<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: Denys Zaruin</title>
    <description>The latest articles on DEV Community by Denys Zaruin (@moveinfra-eu).</description>
    <link>https://dev.to/moveinfra-eu</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4214821%2F878fb0ed-6f24-4e67-b55e-7a869abc64b1.png</url>
      <title>DEV Community: Denys Zaruin</title>
      <link>https://dev.to/moveinfra-eu</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/moveinfra-eu"/>
    <language>en</language>
    <item>
      <title>Is an AWS region in Frankfurt enough for GDPR and data sovereignty?</title>
      <dc:creator>Denys Zaruin</dc:creator>
      <pubDate>Tue, 17 Nov 2026 15:21:58 +0000</pubDate>
      <link>https://dev.to/moveinfra-eu/is-an-aws-region-in-frankfurt-enough-for-gdpr-and-data-sovereignty-3h2f</link>
      <guid>https://dev.to/moveinfra-eu/is-an-aws-region-in-frankfurt-enough-for-gdpr-and-data-sovereignty-3h2f</guid>
      <description>&lt;p&gt;&lt;em&gt;Originally published on &lt;a href="https://moveinfra.eu/blog/is-aws-frankfurt-enough-for-gdpr/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;Yes, an AWS region in Frankfurt (&lt;code&gt;eu-central-1&lt;/code&gt;) is usually enough for GDPR compliance when paired with the right contracts and technical safeguards, because the GDPR does not strictly require personal data to stay inside the European Union. However, hosting in Frankfurt does not give you European data sovereignty. Selecting a German data center gives you data residency, meaning your files sit on servers in Hesse, but the infrastructure is still ultimately owned and controlled by Amazon, a US corporation subject to extraterritorial US laws such as the CLOUD Act. &lt;em&gt;Note: This article is for informational purposes for technical and business leaders and is not legal advice.&lt;/em&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  Key takeaways
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Data residency, GDPR compliance, and data sovereignty are three different things.&lt;/strong&gt; Putting servers in Frankfurt solves residency and helps with GDPR, but does not change which country's laws govern the cloud provider.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;GDPR is about protection, not just geography.&lt;/strong&gt; Under Chapter V of the GDPR, personal data can be processed by US-owned providers or transferred outside the EEA if valid legal mechanisms—such as the EU-US Data Privacy Framework (DPF) or Standard Contractual Clauses (SCCs)—are in place.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;The EU-US Data Privacy Framework is valid as of 2026, but under legal scrutiny.&lt;/strong&gt; Following the EU General Court's dismissal of the &lt;em&gt;Latombe&lt;/em&gt; challenge in September 2025, an appeal (&lt;code&gt;Case C-703/25 P&lt;/code&gt;) is pending before the European Court of Justice, alongside questions over the independence of US oversight bodies.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;US law follows the parent company, not the server rack.&lt;/strong&gt; Under the US CLOUD Act, US authorities can compel a US service provider to produce data in its "possession, custody, or control," regardless of whether that data is stored in Virginia, Frankfurt, or the AWS European Sovereign Cloud.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Not every company needs an EU-owned cloud.&lt;/strong&gt; Standard B2B SaaS startups often do fine in AWS Frankfurt, while companies selling to European public sector bodies, healthcare institutions, defense, or strict financial enterprises increasingly benefit from EU-owned infrastructure.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Three terms that get mixed up: residency, compliance, and sovereignty
&lt;/h2&gt;

&lt;p&gt;When European customers or auditors ask where your data is hosted, "AWS Frankfurt" is the most common answer. To know whether that answer is sufficient for your business, you have to separate three concepts that vendors often blur together.&lt;/p&gt;

&lt;h3&gt;
  
  
  1. Data residency (Where the disks live)
&lt;/h3&gt;

&lt;p&gt;Data residency is simply the physical and geographic location where your data is stored and processed. When you deploy an RDS database or an S3 bucket in AWS &lt;code&gt;eu-central-1&lt;/code&gt; (Frankfurt), AWS commits not to move your customer content out of that region without your consent or instruction.&lt;/p&gt;

&lt;p&gt;Residency solves latency for Central European users and satisfies basic contractual clauses that state "Customer data must be stored within the European Economic Area (EEA)."&lt;/p&gt;

&lt;h3&gt;
  
  
  2. GDPR compliance (How personal data is protected)
&lt;/h3&gt;

&lt;p&gt;The General Data Protection Regulation (Regulation (EU) 2016/679) does not contain a blanket ban on non-EU cloud providers, nor does it mandate that all data remain inside EU borders. Instead, GDPR requires that:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;You have a lawful basis to process personal data and a Data Processing Addendum (DPA) with your processor under Article 28.&lt;/li&gt;
&lt;li&gt;You implement appropriate technical and organizational security measures under Article 32.&lt;/li&gt;
&lt;li&gt;If personal data is transferred outside the EEA—including remote access by support engineers in Seattle or Bangalore, or routing through global control planes and billing systems—that transfer must comply with Chapter V of the GDPR (Articles 44 to 49).&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;Because AWS offers a GDPR-compliant DPA, keeps application data in Frankfurt, and certifies under the EU-US Data Privacy Framework, using AWS Frankfurt can be GDPR compliant for most commercial workloads.&lt;/p&gt;

&lt;h3&gt;
  
  
  3. Data sovereignty (Who has ultimate legal authority)
&lt;/h3&gt;

&lt;p&gt;Data sovereignty is about jurisdiction and control. It asks a blunt question: if a foreign government demands access to your data, or orders your provider to suspend your account, whose law wins?&lt;/p&gt;

&lt;p&gt;True European sovereignty means the cloud provider is headquartered in Europe, incorporated under European law, operated by staff in Europe, and free from parent companies in third countries whose laws claim extraterritorial reach. An American hyperscaler operating a data center in Germany provides European data residency, not European sovereignty.&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Requirement&lt;/th&gt;
&lt;th&gt;AWS Standard Region (&lt;code&gt;eu-central-1&lt;/code&gt; Frankfurt)&lt;/th&gt;
&lt;th&gt;AWS European Sovereign Cloud (Brandenburg)&lt;/th&gt;
&lt;th&gt;EU-Owned Cloud (e.g., UpCloud, Scaleway, OVHcloud, Hetzner)&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Physical server location&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Germany (EU)&lt;/td&gt;
&lt;td&gt;Germany (EU)&lt;/td&gt;
&lt;td&gt;EU Member States&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Control plane &amp;amp; metadata&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Integrated with global AWS network&lt;/td&gt;
&lt;td&gt;Logically isolated in the EU&lt;/td&gt;
&lt;td&gt;Inside the EU&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Operations &amp;amp; support staff&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Global workforce&lt;/td&gt;
&lt;td&gt;EU residents/citizens only&lt;/td&gt;
&lt;td&gt;Primarily EU-based workforce&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Ultimate parent company&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Amazon.com, Inc. (US)&lt;/td&gt;
&lt;td&gt;Amazon.com, Inc. (US)&lt;/td&gt;
&lt;td&gt;European parent company (Finland, France, Germany)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Subject to US CLOUD Act?&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Yes&lt;/td&gt;
&lt;td&gt;Yes (via US ownership)&lt;/td&gt;
&lt;td&gt;No (assuming no considerable US nexus)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Typical cost level&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Standard hyperscaler pricing&lt;/td&gt;
&lt;td&gt;Premium over standard AWS&lt;/td&gt;
&lt;td&gt;Usually 30% to 60% lower on compute and egress&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;h2&gt;
  
  
  The jurisdiction gap: Why Frankfurt is still within reach of US law
&lt;/h2&gt;

&lt;p&gt;The core tension between European privacy law and US cloud providers comes down to a conflict of laws.&lt;/p&gt;

&lt;p&gt;In 2018, the United States enacted the Clarifying Lawful Overseas Use of Data Act (CLOUD Act), which amended the Stored Communications Act (18 U.S.C. § 2713). The statute states that a provider of electronic communication or remote computing services must comply with valid US obligations to preserve, backup, or disclose the contents of a wire or electronic communication and any record or other information pertaining to a customer within the provider's &lt;strong&gt;"possession, custody, or control,"&lt;/strong&gt; regardless of whether that communication or record is located within or outside of the United States.&lt;/p&gt;

&lt;p&gt;At the same time, Article 48 of the GDPR states that any judgment of a court or tribunal and any decision of an administrative authority of a third country requiring a controller or processor to transfer or disclose personal data may only be recognized or enforceable if based on an international agreement, such as a mutual legal assistance treaty (MLAT).&lt;/p&gt;

&lt;p&gt;In daily practice, AWS publishes transparency reports showing that it challenges overbroad requests, redirects government agencies to contact the enterprise customer directly, and discloses enterprise content outside the US only in rare cases. If your threat model is routine commercial risk, the statistical probability of a US law enforcement warrant targeting your Frankfurt PostgreSQL database is very low.&lt;/p&gt;

&lt;p&gt;However, for European public sector buyers, banks, healthcare providers, and defense contractors, procurement decisions are based on legal exposure rather than historical averages. If a US court orders a US parent company to hand over data or restrict service, a subsidiary in Luxembourg or Germany faces a direct conflict between US federal law and EU law.&lt;/p&gt;

&lt;h3&gt;
  
  
  What about the AWS European Sovereign Cloud?
&lt;/h3&gt;

&lt;p&gt;In January 2026, AWS launched the AWS European Sovereign Cloud in Brandenburg, Germany. It is a separate partition from standard AWS regions, with its own billing, identity management, and European certificate authority, operated exclusively by EU residents under a dedicated European corporate structure.&lt;/p&gt;

&lt;p&gt;This solves two practical problems that standard AWS Frankfurt had:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;Metadata and control planes:&lt;/strong&gt; In standard regions, account metadata, IAM configurations, and billing records can touch global systems. In the Sovereign Cloud, metadata stays in the EU.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Operational autonomy:&lt;/strong&gt; US-based engineers have no access to the Sovereign Cloud partition, reducing the risk that day-to-day support triggers a third-country data transfer.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;Yet two caveats remain. First, because it is a separate partition, moving from &lt;code&gt;eu-central-1&lt;/code&gt; to the AWS European Sovereign Cloud requires migrating your accounts and workloads, and the sovereign region carries a price premium over standard Frankfurt pricing. Second, the ultimate owner of the infrastructure remains Amazon.com, Inc., a US company. While operational isolation makes direct data extraction without EU staff involvement technically difficult, it does not eliminate the legal jurisdiction of the United States over the parent corporation.&lt;/p&gt;

&lt;h2&gt;
  
  
  The EU-US Data Privacy Framework in 2026: Valid, but fragile
&lt;/h2&gt;

&lt;p&gt;Even when you host in AWS Frankfurt, using a US hyperscaler almost always involves some international data transfer mechanisms—whether for enterprise support cases, global CDN routing, or third-party sub-processors.&lt;/p&gt;

&lt;p&gt;Historically, European companies relied on Safe Harbor (struck down by the European Court of Justice in &lt;em&gt;Schrems I&lt;/em&gt; in 2015) and then Privacy Shield (struck down in &lt;em&gt;Schrems II&lt;/em&gt; in July 2020). In July 2023, the European Commission adopted its third adequacy decision, creating the &lt;strong&gt;EU-US Data Privacy Framework (DPF)&lt;/strong&gt; (Implementing Decision (EU) 2023/1795), backed by US Executive Order 14086.&lt;/p&gt;

&lt;p&gt;As of late 2026, here is where the legal foundation stands:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;The DPF is currently in force:&lt;/strong&gt; AWS, Google Cloud, and Microsoft Azure are all certified under the DPF. Under Article 45 of the GDPR, transfers to certified US organizations do not require additional transfer tools.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Pending appeal at the CJEU:&lt;/strong&gt; French MP Philippe Latombe challenged the DPF before the EU General Court (&lt;code&gt;Case T-553/23&lt;/code&gt;). On September 3, 2025, the General Court dismissed the challenge, ruling that at the time of its decision, the US framework provided essentially equivalent safeguards. However, that ruling was appealed to the Court of Justice of the European Union on October 31, 2025 (&lt;code&gt;Case C-703/25 P&lt;/code&gt;), where it remains pending.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Questions over US independent oversight:&lt;/strong&gt; The 2023 adequacy decision relied heavily on the independence of US oversight bodies, including the Privacy and Civil Liberties Oversight Board (PCLOB), the Data Protection Review Court (DPRC), and the Federal Trade Commission (FTC). Staffing dismissals at the PCLOB in 2025 and the June 2026 US Supreme Court ruling in &lt;em&gt;Trump v. Slaughter&lt;/em&gt; (which curtailed statutory removal protections for independent agency commissioners) have raised fresh questions from privacy advocates and data protection authorities about whether the US political safeguards underpinning the DPF still hold as originally assessed by the European Commission.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Because no one can guarantee whether the DPF will survive the next three to five years without a "Schrems III" invalidation, legal teams continue to keep Standard Contractual Clauses (SCCs) and Transfer Impact Assessments (TIAs) in place as a fallback.&lt;/p&gt;

&lt;h2&gt;
  
  
  Technical safeguards if you stay on AWS Frankfurt
&lt;/h2&gt;

&lt;p&gt;If you decide to keep your infrastructure in AWS Frankfurt, location alone should not be your only defense. Following recommendations issued by the European Data Protection Board (EDPB) after &lt;em&gt;Schrems II&lt;/em&gt;, you can strengthen your GDPR posture with technical and organizational measures:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Verify your DPA and SCCs:&lt;/strong&gt; Ensure your AWS account has the current AWS GDPR Data Processing Addendum active, which includes Standard Contractual Clauses as a backstop if the DPF is ever invalidated.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Lock down regions with AWS Organizations:&lt;/strong&gt; Apply Service Control Policies (SCPs) to block the creation of resources outside EU regions (&lt;code&gt;eu-central-1&lt;/code&gt;, &lt;code&gt;eu-west-1&lt;/code&gt;, &lt;code&gt;eu-north-1&lt;/code&gt;, etc.) so engineers cannot accidentally spin up a test database in &lt;code&gt;us-east-1&lt;/code&gt;.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Keep support and logging local:&lt;/strong&gt; Avoid putting cleartext personal identifiable information (PII) into AWS Support tickets, IAM role names, or resource tags, since global support staff and billing systems may read them.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Evaluate your encryption key model:&lt;/strong&gt; Standard AWS KMS keys are managed within AWS infrastructure. If you require strict technical protection against provider access, the EDPB emphasizes encryption where keys are held solely under the customer's control in the EEA (such as an external HSM or client-side encryption before upload)—though this comes with significant operational complexity and breaks many managed cloud features.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Checklist: Is AWS Frankfurt enough, or is an EU-owned cloud worth it?
&lt;/h2&gt;

&lt;p&gt;Moving clouds takes engineering effort, even when automated. For some companies, staying in Frankfurt is the pragmatic choice; for others, moving to an EU-owned provider (such as UpCloud, Scaleway, OVHcloud, Hetzner, IONOS, or STACKIT) directly removes commercial and regulatory friction.&lt;/p&gt;

&lt;h3&gt;
  
  
  Staying on AWS Frankfurt is usually fine when:
&lt;/h3&gt;

&lt;ul&gt;
&lt;li&gt;[ ] &lt;strong&gt;You build standard B2B or B2C software&lt;/strong&gt; and your customers' security questionnaires only ask that data is hosted in an EU region with a valid DPA.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;You rely heavily on proprietary AWS services&lt;/strong&gt;—such as DynamoDB, Redshift, or complex Step Functions—where refactoring would take months away from core product development.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;You have an active AWS promotional credit grant&lt;/strong&gt; covering your infrastructure bill for the next 12 months.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;You are not subject to strict sector regulations&lt;/strong&gt; regarding third-country cloud concentration or public procurement rules.&lt;/li&gt;
&lt;/ul&gt;

&lt;h3&gt;
  
  
  Moving to an EU-owned cloud is worth it when:
&lt;/h3&gt;

&lt;ul&gt;
&lt;li&gt;[ ] &lt;strong&gt;You sell to the European public sector, healthcare, or defense.&lt;/strong&gt; Public tenders in Germany, France, the Nordics, and the Benelux increasingly award points for—or explicitly require—providers free from non-EU extraterritorial laws.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Enterprise procurement deals are stalling in legal review.&lt;/strong&gt; If your sales cycle drags out by months because a German Mittelstand customer or French bank demands a 40-page Transfer Impact Assessment for every US sub-processor, using an EU-owned host cuts the Chapter V transfer debate out of the conversation.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;You fall under DORA (Regulation (EU) 2022/2554) or NIS2.&lt;/strong&gt; Since January 17, 2025, DORA requires financial entities and their critical ICT providers to maintain tested exit strategies and manage third-country concentration risks.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Your architecture uses standard building blocks.&lt;/strong&gt; If your stack runs on Linux VMs, Kubernetes, PostgreSQL/MySQL, Redis, and S3-compatible object storage, every component has a direct drop-in equivalent on EU clouds (you can check service parity on our free &lt;a href="https://moveinfra.eu/eu-cloud-map/" rel="noopener noreferrer"&gt;EU Cloud Service Map&lt;/a&gt;).&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;You want to cut your monthly cloud bill alongside your legal risk.&lt;/strong&gt; Standard compute, managed databases, and bandwidth on EU-owned providers are substantially cheaper than AWS Frankfurt, and under Article 29 of the EU Data Act (Regulation (EU) 2023/2854), data egress charges for switching cloud providers are capped at cost until January 12, 2027, and banned entirely from January 12, 2027 onward. You can see what your current setup would cost on European providers using our browser-based &lt;a href="https://moveinfra.eu/bill-analyzer/" rel="noopener noreferrer"&gt;Cloud Bill Analyzer&lt;/a&gt;, or read how exit fee waivers work in our &lt;a href="https://moveinfra.eu/eu-data-act-switching-fees/" rel="noopener noreferrer"&gt;EU Data Act switching fees guide&lt;/a&gt;.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Frequently asked questions
&lt;/h2&gt;

&lt;h3&gt;
  
  
  Does GDPR require all personal data to stay inside the EU?
&lt;/h3&gt;

&lt;p&gt;No. GDPR governs the protection of personal data belonging to individuals in the EEA, not its mandatory physical localization. Data can be processed outside the EEA—or by non-EU companies inside the EEA—provided you comply with the conditions in Chapter V of the GDPR, such as relying on an European Commission adequacy decision or Standard Contractual Clauses with appropriate safeguards.&lt;/p&gt;

&lt;h3&gt;
  
  
  Does encrypting my data with AWS KMS in Frankfurt protect me from the US CLOUD Act?
&lt;/h3&gt;

&lt;p&gt;Standard server-side encryption using AWS Key Management Service (SSE-KMS) protects against physical disk theft and unauthorized external access, and AWS provides strong architectural controls around key usage. However, because AWS manages the KMS service, keys managed inside the cloud environment are generally not considered by EU data protection authorities to be under the &lt;em&gt;exclusive&lt;/em&gt; control of the customer unless you manage keys externally (Hold Your Own Key) and encrypt data before it reaches the cloud provider.&lt;/p&gt;

&lt;h3&gt;
  
  
  Is the AWS European Sovereign Cloud immune to the US CLOUD Act?
&lt;/h3&gt;

&lt;p&gt;The AWS European Sovereign Cloud provides strong operational isolation: it is physically and logically separate from global AWS regions, keeps metadata in Europe, and is operated solely by EU residents. However, because the infrastructure is ultimately owned by Amazon.com, Inc., a US corporation, legal scholars and European regulators note that it does not completely eliminate the extraterritorial reach of US law over the parent company.&lt;/p&gt;

&lt;h3&gt;
  
  
  How hard is it to move from AWS Frankfurt to an EU-owned cloud?
&lt;/h3&gt;

&lt;p&gt;If your application relies on open-source standards—such as Docker or Kubernetes, PostgreSQL or MySQL, Redis, and S3-compatible object storage—moving is primarily an infrastructure-as-code and database replication task rather than an application rewrite. Under the EU Data Act and AWS's own switching policies, data egress fees for leaving the platform can also be waived when following formal exit procedures.&lt;/p&gt;

&lt;h2&gt;
  
  
  Sources
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;&lt;a href="https://eur-lex.europa.eu/eli/reg/2016/679/oj" rel="noopener noreferrer"&gt;Regulation (EU) 2016/679 (General Data Protection Regulation) — EUR-Lex&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://eur-lex.europa.eu/eli/dec_impl/2023/1795/oj" rel="noopener noreferrer"&gt;Commission Implementing Decision (EU) 2023/1795 on the adequate level of protection of personal data under the EU-US Data Privacy Framework — EUR-Lex&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://eur-lex.europa.eu/eli/reg/2023/2854/oj" rel="noopener noreferrer"&gt;Regulation (EU) 2023/2854 (EU Data Act) — EUR-Lex&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://www.dataprivacyframework.gov/" rel="noopener noreferrer"&gt;EU-U.S. Data Privacy Framework Program — Official Participant List and Overview&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://aws.amazon.com/compliance/gdpr-center/" rel="noopener noreferrer"&gt;AWS GDPR Center — Amazon Web Services Compliance&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;




&lt;p&gt;&lt;em&gt;Drafted with AI assistance and fact-checked by the MoveInfra team. The original, kept up to date, is at &lt;a href="https://moveinfra.eu/blog/is-aws-frankfurt-enough-for-gdpr/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu&lt;/a&gt;. Free tools for pricing a move to an EU cloud: &lt;a href="https://moveinfra.eu/tools/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu/tools&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>cloud</category>
      <category>gdpr</category>
      <category>datasovereignty</category>
      <category>aws</category>
    </item>
    <item>
      <title>How to estimate savings from moving your AWS bill to an EU cloud</title>
      <dc:creator>Denys Zaruin</dc:creator>
      <pubDate>Wed, 11 Nov 2026 13:21:39 +0000</pubDate>
      <link>https://dev.to/moveinfra-eu/how-to-estimate-savings-from-moving-your-aws-bill-to-an-eu-cloud-2dlj</link>
      <guid>https://dev.to/moveinfra-eu/how-to-estimate-savings-from-moving-your-aws-bill-to-an-eu-cloud-2dlj</guid>
      <description>&lt;p&gt;&lt;em&gt;Originally published on &lt;a href="https://moveinfra.eu/blog/estimate-savings-from-aws-bill/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;To estimate how much you can save by moving from AWS to a European cloud provider, export your detailed AWS invoice, group the line items into core categories (compute, storage, database, network), map each resource to its equivalent by vCPU, RAM, and gigabytes, and compare your net discounted AWS spend against European public list prices. If you want an immediate estimate without building a spreadsheet, you can upload your AWS bill export to our free &lt;a href="https://moveinfra.eu/bill-analyzer/" rel="noopener noreferrer"&gt;Bill Analyzer&lt;/a&gt;, which calculates the equivalent monthly cost on UpCloud, Hetzner, Scaleway, and OVHcloud directly in your browser so the file never leaves your device.&lt;/p&gt;

&lt;h2&gt;
  
  
  Key takeaways
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Look past headline compute rates:&lt;/strong&gt; While virtual machines on European clouds are often lower in price than AWS on-demand instances, the largest cost reductions frequently come from eliminating internet egress and NAT gateway fees.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Compare net spend, not list prices:&lt;/strong&gt; If your company uses AWS Savings Plans, Reserved Instances, or Enterprise Discount Programs, compare the European provider's price against your actual discounted AWS bill, not AWS on-demand rates.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Account for missing managed services:&lt;/strong&gt; Some lower-cost providers do not offer managed databases or managed Kubernetes. You must factor in the engineering time or extra virtual machines required to run those services yourself.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Savings vary widely by architecture:&lt;/strong&gt; Standard web applications with high outbound traffic usually see substantial savings, while workloads tied to proprietary services like DynamoDB or BigQuery require refactoring that changes the financial calculation.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Why AWS bills are difficult to compare by hand
&lt;/h2&gt;

&lt;p&gt;An AWS invoice rarely gives you a clean list of servers and hard drives. Instead, a single application might generate dozens of billing lines across EC2, EBS, RDS, S3, Elastic Load Balancing, and Data Transfer. &lt;/p&gt;

&lt;p&gt;AWS also separates the cost of a resource from the cost of using it. When you run a virtual machine on EC2, you pay for the compute hours, the attached EBS boot volume, the provisioned IOPS if you exceed the baseline, the public IPv4 address, the NAT gateway processing fees, and every gigabyte of data sent out to the internet. &lt;/p&gt;

&lt;p&gt;European infrastructure providers—such as UpCloud, Scaleway, OVHcloud, and Hetzner—generally bundle more of these elements together. Bandwidth allowances are frequently included with the server, and network architectures tend to require fewer billable middleboxes. Because the billing structures differ, estimating your savings requires translating your AWS usage metrics into raw hardware and bandwidth specifications before looking at prices.&lt;/p&gt;

&lt;h2&gt;
  
  
  A 6-step method to estimate your European cloud cost
&lt;/h2&gt;

&lt;p&gt;If you are building a business case for a CFO or engineering lead, a back-of-the-envelope guess is not enough. Follow these six steps to build a defensible cost comparison.&lt;/p&gt;

&lt;h3&gt;
  
  
  Step 1: Export a detailed month of AWS billing data
&lt;/h3&gt;

&lt;p&gt;Do not use the high-level summary on the first page of your AWS PDF invoice. It groups costs by service name without showing instance sizes or data volumes.&lt;/p&gt;

&lt;p&gt;Go to the AWS Billing and Cost Management console and open &lt;strong&gt;Bills&lt;/strong&gt;. Select the most recent full calendar month that represents normal traffic for your business (avoid seasonal peaks or months where you ran one-off data migrations). Expand the &lt;strong&gt;Charges by service&lt;/strong&gt; section to see the regional breakdown and exact usage quantities—such as &lt;code&gt;BoxUsage:m6a.xlarge&lt;/code&gt; measured in hours, or &lt;code&gt;DataTransfer-Out-Bytes&lt;/code&gt; measured in gigabytes. You can also export a CSV from AWS Cost Explorer grouped by &lt;strong&gt;Usage Type&lt;/strong&gt; and &lt;strong&gt;Service&lt;/strong&gt;.&lt;/p&gt;

&lt;p&gt;For a faster alternative, you can drop your billing export into our free &lt;a href="https://moveinfra.eu/bill-analyzer/" rel="noopener noreferrer"&gt;Bill Analyzer&lt;/a&gt;. It parses the usage quantities locally in your browser—meaning your billing data is never uploaded to a server—and maps them to European cloud prices automatically.&lt;/p&gt;

&lt;h3&gt;
  
  
  Step 2: Group line items into five core buckets
&lt;/h3&gt;

&lt;p&gt;Filter out non-infrastructure costs like AWS Marketplace software subscriptions, domain registrations, and support plans (we will deal with support in Step 5). Group the remaining technical line items into five buckets:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;Compute:&lt;/strong&gt; EC2 instances, Fargate vCPU/memory hours, and EKS control plane fees.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Databases and caches:&lt;/strong&gt; RDS instances, ElastiCache nodes, and database storage.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Storage:&lt;/strong&gt; EBS volumes, EBS snapshots, and S3 buckets.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Networking and load balancing:&lt;/strong&gt; Application Load Balancers (ALB), Network Load Balancers (NLB), NAT Gateways, public IPv4 addresses, and all Data Transfer Out lines.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Cloud-native and proprietary services:&lt;/strong&gt; SQS, SNS, Lambda, DynamoDB, Redshift, CloudWatch.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;Convert hourly figures into monthly counts. In cloud billing, a standard month is 730 hours. If an EC2 usage line shows 2,920 hours of &lt;code&gt;m6a.xlarge&lt;/code&gt; in a month, divide 2,920 by 730 to find that you are running four continuous instances.&lt;/p&gt;

&lt;h3&gt;
  
  
  Step 3: Map AWS resources to European equivalents by specification
&lt;/h3&gt;

&lt;p&gt;Never map cloud servers by marketing tier names. An "extra large" or "general purpose" instance means different things on different platforms. Map every resource strictly by its underlying hardware specification:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;vCPU count and architecture:&lt;/strong&gt; Check whether your AWS instances use x86 (Intel/AMD, such as &lt;code&gt;m6i&lt;/code&gt; or &lt;code&gt;m6a&lt;/code&gt;) or ARM (&lt;code&gt;m7g&lt;/code&gt; Graviton). Match the exact number of vCPUs.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Memory (RAM):&lt;/strong&gt; Match the gigabytes of RAM. If a European provider does not have an exact match (for example, offering 4 vCPU with 8 GB or 16 GB, while your workload uses 12 GB), always round up to the next size to avoid memory starvation.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;CPU allocation (Shared vs. Dedicated):&lt;/strong&gt; AWS &lt;code&gt;t3&lt;/code&gt; and &lt;code&gt;t4g&lt;/code&gt; instances use burstable, shared CPUs, whereas &lt;code&gt;m6a&lt;/code&gt; and &lt;code&gt;c6i&lt;/code&gt; use dedicated resources. When pricing European clouds, compare AWS &lt;code&gt;m&lt;/code&gt; or &lt;code&gt;c&lt;/code&gt; series against dedicated vCPU plans (like Hetzner CCX, Scaleway PRO2/GP1, OVHcloud b3, or UpCloud Premium), not against cheap shared-CPU starter plans.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;You can use our &lt;a href="https://moveinfra.eu/eu-cloud-map/" rel="noopener noreferrer"&gt;EU Cloud Map&lt;/a&gt; to see which European services correspond directly to AWS compute, storage, and managed database tiers.&lt;/p&gt;

&lt;h3&gt;
  
  
  Step 4: Apply public list prices
&lt;/h3&gt;

&lt;p&gt;Once you have a clean bill of materials (for example: four 8 vCPU / 32 GB servers, one 2 vCPU / 8 GB managed PostgreSQL database, 2 TB of block storage, and 10 TB of egress), apply the target provider's list prices.&lt;/p&gt;

&lt;p&gt;The table below shows approximate public list prices as of 2026 (fetched 28 October 2026 from our open-data &lt;a href="https://moveinfra.eu/eu-cloud-prices/" rel="noopener noreferrer"&gt;EU Cloud Price Index&lt;/a&gt;). All figures are in EUR per month, excluding VAT, with no discounts or commitments, converting USD at the ECB rate of 1.1133. A month is calculated at 730 hours or the provider's monthly cap (672 hours for UpCloud), comparing AWS Frankfurt (&lt;code&gt;eu-central-1&lt;/code&gt;) against European providers. Always check each provider's official pricing page for current rates before finalizing your budget.&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Resource Category&lt;/th&gt;
&lt;th&gt;AWS (Frankfurt)&lt;/th&gt;
&lt;th&gt;Hetzner&lt;/th&gt;
&lt;th&gt;OVHcloud&lt;/th&gt;
&lt;th&gt;Scaleway&lt;/th&gt;
&lt;th&gt;UpCloud&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Server: 2 vCPU, 8 GB&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€67.86 (&lt;code&gt;m6a.large&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€42.99 (&lt;code&gt;CCX13&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€37.38 (&lt;code&gt;b3-8&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€40.95 (&lt;code&gt;PRO2-XXS&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€38.00 (&lt;code&gt;PREMIUM-2xCPU-8GB&lt;/code&gt;)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Server: 4 vCPU, 16 GB&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€135.73 (&lt;code&gt;m6a.xlarge&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€85.99 (&lt;code&gt;CCX23&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€74.68 (&lt;code&gt;b3-16&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€67.76 (&lt;code&gt;GP1-XS&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€92.00 (&lt;code&gt;PREMIUM-4xCPU-16GB&lt;/code&gt;)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Server: 8 vCPU, 32 GB&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€271.46 (&lt;code&gt;m6a.2xlarge&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€138.49 (&lt;code&gt;CCX33&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€149.36 (&lt;code&gt;b3-32&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€139.24 (&lt;code&gt;GP1-S&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€192.00 (&lt;code&gt;PREMIUM-8xCPU-32GB&lt;/code&gt;)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Server: 16 vCPU, 64 GB&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€542.93 (&lt;code&gt;m6a.4xlarge&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€275.99 (&lt;code&gt;CCX43&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€298.72 (&lt;code&gt;b3-64&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€279.97 (&lt;code&gt;GP1-M&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€384.00 (&lt;code&gt;PREMIUM-16xCPU-64GB&lt;/code&gt;)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Block storage: 1 TB SSD&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€85.50 (&lt;code&gt;EBS gp3&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€44.00 (&lt;code&gt;Volumes&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€86.00 (&lt;code&gt;High-speed&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€94.90 (&lt;code&gt;5K&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€220.00 (&lt;code&gt;MaxIOPS&lt;/code&gt;)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Object storage: 1 TB&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€22.00 (&lt;code&gt;S3 Standard&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€6.40 (&lt;code&gt;Object Storage&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€14.00 (&lt;code&gt;Standard 3-AZ&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€16.10 (&lt;code&gt;Multi-AZ&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€20.00 (&lt;code&gt;Managed&lt;/code&gt;)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Managed Load Balancer&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€17.70 (&lt;code&gt;ALB&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€7.49 (&lt;code&gt;LB11&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€6.00 (&lt;code&gt;Octavia Small&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€16.79 (&lt;code&gt;LB-S&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€30.00 (&lt;code&gt;Production S&lt;/code&gt;)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Managed K8s control plane&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€65.57 (&lt;code&gt;EKS Standard&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;&lt;em&gt;No managed offer&lt;/em&gt;&lt;/td&gt;
&lt;td&gt;€65.70 (&lt;code&gt;MKS Standard&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€80.30 (&lt;code&gt;Dedicated-4&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€60.00 (&lt;code&gt;Production&lt;/code&gt;)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Managed PostgreSQL: 2 vCPU, 8 GB&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€131.14 (&lt;code&gt;db.m7g.large&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;&lt;em&gt;No managed offer&lt;/em&gt;&lt;/td&gt;
&lt;td&gt;€143.15 (&lt;code&gt;Essential B3-8&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€80.30 (&lt;code&gt;DB-PRO2-XXS&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€60.00 (&lt;code&gt;rdb.standard&lt;/code&gt;)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Internet egress: 1 TB (1 server)&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€72.72&lt;/td&gt;
&lt;td&gt;€0.00*&lt;/td&gt;
&lt;td&gt;€0.00*&lt;/td&gt;
&lt;td&gt;€0.00*&lt;/td&gt;
&lt;td&gt;€0.00*&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Internet egress: 10 TB (1 server)&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€799.92&lt;/td&gt;
&lt;td&gt;€0.00*&lt;/td&gt;
&lt;td&gt;€0.00*&lt;/td&gt;
&lt;td&gt;€0.00*&lt;/td&gt;
&lt;td&gt;€0.00*&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;&lt;em&gt;*Note on €0.00 egress: On these European providers, €0.00 means there is no excess bandwidth fee for 1 TB or 10 TB within the included traffic allowances and fair-use terms of their standard servers. Always check each provider's specific bandwidth limits and port speed terms.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;As the table shows, price dynamics vary by component:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Compute and Egress:&lt;/strong&gt; Dedicated virtual machines and outbound internet traffic are consistently less expensive on European clouds than AWS on-demand list prices. For example, 10 TB of internet egress from a single server costs approximately €799.92 on AWS as of 2026, while falling within the included €0.00 traffic allowance on Hetzner, OVHcloud, Scaleway, and UpCloud.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Storage and Managed Services:&lt;/strong&gt; High-performance block storage on UpCloud (€220.00 for 1 TB MaxIOPS) or Scaleway (€94.90 for 1 TB 5K) is higher than AWS EBS &lt;code&gt;gp3&lt;/code&gt; base capacity (€85.50). Similarly, OVHcloud's Essential B3-8 managed PostgreSQL (€143.15) is slightly higher than an AWS &lt;code&gt;db.m7g.large&lt;/code&gt; single-AZ instance (€131.14). You must price the entire stack together rather than cherry-picking individual lines.&lt;/li&gt;
&lt;/ul&gt;

&lt;h3&gt;
  
  
  Step 5: Add what is missing (Self-managed overhead, HA, and support)
&lt;/h3&gt;

&lt;p&gt;A common mistake in migration estimates is comparing a highly available, fully managed AWS setup against a bare-bones virtual machine setup. Adjust your European cloud estimate for three factors:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;Unmatched managed services:&lt;/strong&gt; Notice in the table above that Hetzner has no managed Kubernetes or managed PostgreSQL offer. If you move an AWS RDS database to Hetzner, you must provision virtual machines, configure replication (such as Patroni), set up automated backups to object storage, and handle patching yourself. Even if the raw server is cheap, your team takes on operational risk and maintenance hours.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;High availability (Multi-AZ):&lt;/strong&gt; Check if your AWS RDS or Load Balancer lines are Multi-AZ. If your AWS database is Multi-AZ, ensure the European equivalent you price includes a standby node, rather than comparing a two-node AWS cluster against a single-node European database.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Support plans:&lt;/strong&gt; AWS charges a percentage of your monthly bill for Developer, Business, or Enterprise Support. Some European providers include basic ticket support for free and charge fixed monthly tiers for 24/7 phone support and SLA response times. Add the appropriate support tier to both sides of your sheet.&lt;/li&gt;
&lt;/ol&gt;

&lt;h3&gt;
  
  
  Step 6: Compare against your net AWS spend
&lt;/h3&gt;

&lt;p&gt;Before calculating the difference, look at the bottom of your AWS bill for &lt;strong&gt;Savings Plans&lt;/strong&gt;, &lt;strong&gt;Reserved Instances&lt;/strong&gt;, or &lt;strong&gt;Enterprise Discount Program (EDP)&lt;/strong&gt; credits. &lt;/p&gt;

&lt;p&gt;If you are comparing European list prices against AWS on-demand prices, your estimate will look artificially high if your company actually pays 25% to 40% less on AWS due to 1-year or 3-year commitments. Conversely, remember that many European providers also offer committed-use discounts or monthly billing caps if you are willing to lock in capacity.&lt;/p&gt;

&lt;p&gt;When we combine a typical web application stack—including compute, a 2 vCPU / 8 GB managed PostgreSQL database, block and object storage, a load balancer, and moderate traffic—our benchmark example stack in the &lt;a href="https://moveinfra.eu/eu-cloud-prices/" rel="noopener noreferrer"&gt;EU Cloud Price Index&lt;/a&gt; totals approximately (as of 28 October 2026):&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;AWS:&lt;/strong&gt; €774.30 / month (on-demand list price)&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;UpCloud:&lt;/strong&gt; €496.00 / month (with managed PostgreSQL)&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;OVHcloud:&lt;/strong&gt; €430.19 / month (with managed PostgreSQL)&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Scaleway:&lt;/strong&gt; €363.92 / month (with managed PostgreSQL)&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Hetzner:&lt;/strong&gt; €336.85 / month (no managed PostgreSQL; self-run on a 2 vCPU, 8 GB server)&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Your own workload will differ depending on your ratio of compute to storage and how much data you send over the internet.&lt;/p&gt;

&lt;h2&gt;
  
  
  Checklist: Common hidden costs in an AWS-to-EU estimate
&lt;/h2&gt;

&lt;p&gt;Use this checklist to verify your spreadsheet before presenting it to leadership:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;[ ] &lt;strong&gt;NAT Gateway processing fees removed?&lt;/strong&gt; On AWS, you pay an hourly fee plus a per-gigabyte processing fee for every gigabyte passing through a NAT Gateway. Most European clouds provide simpler private networking or free virtual routers where these per-gigabyte internal routing charges do not exist.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Cross-AZ traffic accounted for?&lt;/strong&gt; AWS charges for data transferred between Availability Zones in the same region (for example, between EC2 nodes and RDS in different AZs). Check whether your target European provider charges for internal private network traffic.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Public IPv4 addresses counted?&lt;/strong&gt; AWS charges approximately $0.005 per hour ($3.65/month) per public IPv4 address. Check if your target provider includes one IPv4 per server or charges a separate monthly fee.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Commitments expiration dates checked?&lt;/strong&gt; If you have active AWS Compute Savings Plans, you will continue paying that hourly commitment to AWS until the term expires, even if you move the servers away tomorrow.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Migration and parallel-run costs budgeted?&lt;/strong&gt; During a migration, you will briefly run infrastructure on both clouds at once to sync databases and test deployments. Under EU Data Act Article 34, providers may still charge at cost for parallel services.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Do not forget switching and egress fees
&lt;/h2&gt;

&lt;p&gt;When calculating the return on investment (ROI) of a cloud migration, you must compare your monthly savings against the one-time cost of moving. Historically, AWS charged steep data egress fees to copy terabytes of S3 buckets and RDS databases out to another provider.&lt;/p&gt;

&lt;p&gt;Under the &lt;strong&gt;EU Data Act&lt;/strong&gt; (Regulation (EU) 2023/2854, Article 29), cloud providers cannot profit from switching fees: data egress charges for switching may be billed at no more than actual cost until 12 January 2027, and switching charges (including egress for switching) are completely banned starting 12 January 2027.&lt;/p&gt;

&lt;p&gt;Furthermore, as of 2026, major clouds operate specific exit waiver programs:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;AWS&lt;/strong&gt; waives data transfer out fees when a customer moves all data off AWS, or all data out of a specific AWS service, provided you contact AWS Support in advance for approval. European customers can also invoke switching terms under the AWS EU Data Act Addendum.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Google Cloud&lt;/strong&gt; and &lt;strong&gt;Azure&lt;/strong&gt; have similar exit credit programs when terminating service agreements (and for comparable stacks in our index, Azure totals €764.76 and Google Cloud €746.58 as of 28 October 2026).&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Always verify the current program terms on AWS's official pricing and legal pages before initiating a transfer. You can estimate your one-time transfer costs and generate a formal switching notice using our free &lt;a href="https://moveinfra.eu/cloud-exit-fees/" rel="noopener noreferrer"&gt;Cloud Exit Fee Calculator&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  Frequently asked questions
&lt;/h2&gt;

&lt;h3&gt;
  
  
  Can I just multiply my AWS bill by a percentage to estimate savings?
&lt;/h3&gt;

&lt;p&gt;No. Anyone promising a flat percentage reduction without looking at your bill is guessing. A media company paying €3,000 a month mostly in AWS CloudFront and EC2 data egress will see a completely different savings profile than a SaaS company paying €3,000 a month for heavily discounted Reserved Instances and DynamoDB tables.&lt;/p&gt;

&lt;h3&gt;
  
  
  How do I price AWS proprietary services like DynamoDB, SQS, or Lambda?
&lt;/h3&gt;

&lt;p&gt;If you use serverless functions or queues at low volumes, the cost on your AWS bill is often just a few euros. However, moving them to a European cloud may require replacing them with open-source equivalents—such as running RabbitMQ or Redis for queues, or moving Lambda functions into containers on Kubernetes or managed container platforms (like Scaleway Serverless Containers). Price the compute required to host those replacements, and factor in the engineering time needed to adapt your code.&lt;/p&gt;

&lt;h3&gt;
  
  
  Why is block storage sometimes more expensive on European clouds than on AWS?
&lt;/h3&gt;

&lt;p&gt;On AWS, the base price of &lt;code&gt;gp3&lt;/code&gt; EBS storage (€85.50 per TB as of 28 October 2026) includes a baseline of 3,000 IOPS and 125 MB/s throughput; if you need higher performance, AWS bills you extra per provisioned IOPS and MB/s. Some European providers, like UpCloud with its MaxIOPS storage (€220.00 per TB), only offer high-performance enterprise storage tiers by default rather than a slower baseline tier.&lt;/p&gt;

&lt;h3&gt;
  
  
  What if we want to stay on AWS and use the AWS European Sovereign Cloud?
&lt;/h3&gt;

&lt;p&gt;The AWS European Sovereign Cloud keeps data and operations inside the European Union, which helps with operational residency requirements, but it still belongs to Amazon (a US jurisdiction company) and uses AWS's pricing model. Moving to the AWS Sovereign Cloud is a compliance decision, not a cost-reduction strategy.&lt;/p&gt;

&lt;h2&gt;
  
  
  Sources
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;&lt;a href="https://eur-lex.europa.eu/eli/reg/2023/2854/oj" rel="noopener noreferrer"&gt;Regulation (EU) 2023/2854 (EU Data Act) on EUR-Lex&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://aws.amazon.com/ec2/pricing/on-demand/" rel="noopener noreferrer"&gt;AWS EC2 On-Demand Pricing&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://upcloud.com/pricing/" rel="noopener noreferrer"&gt;UpCloud Pricing&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://www.scaleway.com/en/pricing/" rel="noopener noreferrer"&gt;Scaleway Pricing&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://www.ovhcloud.com/en-ie/public-cloud/prices/" rel="noopener noreferrer"&gt;OVHcloud Public Cloud Pricing&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://www.hetzner.com/cloud" rel="noopener noreferrer"&gt;Hetzner Cloud Pricing&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;




&lt;p&gt;&lt;em&gt;Drafted with AI assistance and fact-checked by the MoveInfra team. The original, kept up to date, is at &lt;a href="https://moveinfra.eu/blog/estimate-savings-from-aws-bill/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu&lt;/a&gt;. Free tools for pricing a move to an EU cloud: &lt;a href="https://moveinfra.eu/tools/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu/tools&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>cloud</category>
      <category>aws</category>
      <category>cloudcosts</category>
      <category>europeancloud</category>
    </item>
    <item>
      <title>How to choose a cloud exit consultancy for a move to an EU cloud</title>
      <dc:creator>Denys Zaruin</dc:creator>
      <pubDate>Tue, 10 Nov 2026 14:04:31 +0000</pubDate>
      <link>https://dev.to/moveinfra-eu/how-to-choose-a-cloud-exit-consultancy-for-a-move-to-an-eu-cloud-5449</link>
      <guid>https://dev.to/moveinfra-eu/how-to-choose-a-cloud-exit-consultancy-for-a-move-to-an-eu-cloud-5449</guid>
      <description>&lt;p&gt;&lt;em&gt;Originally published on &lt;a href="https://moveinfra.eu/blog/how-to-choose-a-cloud-exit-consultancy/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;To choose a cloud exit consultancy for a move from AWS to a European-owned cloud, look for a partner that delivers clean Infrastructure as Code (such as Terraform), uses live database replication rather than dump-and-restore scripts, provides a tested rollback mechanism, and works under a fixed or milestone-based contract after an initial discovery phase. Avoid general IT agencies that only hold US hyperscaler certifications. Your contract should explicitly require an audited resource inventory before execution starts, full ownership of all provisioning code, and strict boundaries around customer data access.&lt;/p&gt;

&lt;h2&gt;
  
  
  Key takeaways
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Separate discovery from execution:&lt;/strong&gt; Never sign an open-ended time-and-materials contract for the entire migration. Pay for a bounded inventory and architecture phase first, or require a fixed price once read-only discovery is complete.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Understand who you are hiring:&lt;/strong&gt; Destination cloud providers offer professional services that know their own platform deeply, while independent migration partners help you compare providers, handle multi-provider architectures, and clean up the source AWS environment.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Demand Infrastructure as Code and live sync:&lt;/strong&gt; A serious partner rebuilds your infrastructure cleanly in Terraform or OpenTofu and uses logical replication for databases.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Put rollback and data rules in the contract:&lt;/strong&gt; Require a documented cutover runbook with a defined write-pause window, reverse data sync for post-cutover rollback, and a Data Processing Agreement (DPA) that minimizes human access to production data.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  What a cloud exit consultancy actually does
&lt;/h2&gt;

&lt;p&gt;Moving off AWS to an EU-owned cloud provider—such as UpCloud, Scaleway, Hetzner, OVHcloud, STACKIT, IONOS, or Exoscale—is a different engineering problem than moving workloads into AWS. When companies move &lt;em&gt;into&lt;/em&gt; a US hyperscaler, they often rely on vendor-subsidized migration programs and proprietary migration agents. Moving &lt;em&gt;out&lt;/em&gt; requires translating AWS-specific constructs into portable, standard equivalents.&lt;/p&gt;

&lt;p&gt;A competent cloud exit partner handles five specific areas of work:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;Discovery and service mapping:&lt;/strong&gt; Auditing your AWS billing exports, IAM roles, networks, and Terraform state to find every active resource, including forgotten snapshots, elastic IPs, and regional data transfer paths. You can preview what this mapping looks like using the free &lt;a href="https://moveinfra.eu/eu-cloud-map/" rel="noopener noreferrer"&gt;EU Cloud Service Map&lt;/a&gt;.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Target architecture and sizing:&lt;/strong&gt; Translating AWS instance families, security groups, and VPC setups into the destination provider's networking and compute model. Where an EU cloud lacks a direct managed equivalent (for example, Hetzner does not offer managed databases or managed Kubernetes), the partner must design how those services will run reliably on virtual servers or across a secondary provider.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Infrastructure as Code (IaC) rebuild:&lt;/strong&gt; Writing clean, modular Terraform or OpenTofu code for the new environment so your team can manage it after the consultancy leaves.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Data synchronization and CI/CD integration:&lt;/strong&gt; Setting up continuous logical replication for relational databases (PostgreSQL, MySQL), syncing S3 buckets to EU S3-compatible object storage with checksum verification, and helping your engineers point existing deployment pipelines to the new servers.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Cutover, rollback, and AWS decommissioning:&lt;/strong&gt; Running the switchover during a planned maintenance window, keeping a reverse path open in case of unexpected application issues, and safely tearing down the old AWS accounts so stray bills do not continue.&lt;/li&gt;
&lt;/ol&gt;

&lt;h2&gt;
  
  
  Independent migration partner vs. destination provider professional services
&lt;/h2&gt;

&lt;p&gt;When you decide to move to a European cloud, you generally have two choices for outside help: the destination cloud provider's in-house professional services team, or an independent migration partner. Both have clear strengths and trade-offs.&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Comparison area&lt;/th&gt;
&lt;th&gt;Independent migration partner&lt;/th&gt;
&lt;th&gt;Destination provider professional services&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Provider selection&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Can benchmark multiple EU clouds (e.g., comparing UpCloud vs. OVHcloud vs. Scaleway) against your workload and budget.&lt;/td&gt;
&lt;td&gt;Will architect exclusively for their own platform, even if a specific workload fits better elsewhere.&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Platform depth&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Broad operational knowledge across AWS and several EU clouds; relies on public APIs and standard tooling.&lt;/td&gt;
&lt;td&gt;Direct access to internal product teams, roadmaps, and deep platform-specific troubleshooting.&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Source cloud (AWS) work&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Actively works inside your AWS environment to untangle dependencies, manage exit notices, and shut down old resources.&lt;/td&gt;
&lt;td&gt;Usually stops at their own boundary; gives you guidance on how to push data in, but rarely touches your AWS account.&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Commercial incentives&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Paid for the migration project itself; motivated to finish the cutover and hand over keys on schedule.&lt;/td&gt;
&lt;td&gt;Often subsidized or discounted to win your recurring monthly cloud spend; scope is usually limited to onboarding architecture.&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Hybrid setups&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Comfortable building setups that span providers (e.g., bare metal on one EU provider, managed database on another).&lt;/td&gt;
&lt;td&gt;Designs single-cloud architectures.&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;If you already know which EU provider you want, have an internal DevOps team to handle the AWS teardown, and mainly need architectural validation on the new platform, the destination provider's solutions architects are often enough. If your engineering team is busy shipping product and you want another team to build the Terraform, wire the database replication, coordinate the cutover, and clean up AWS, look for an independent specialist.&lt;/p&gt;

&lt;h2&gt;
  
  
  Questions to ask before you sign
&lt;/h2&gt;

&lt;p&gt;When interviewing potential partners, skip generic questions about how long they have been in business. Ask operational questions that reveal how they actually move production systems.&lt;/p&gt;

&lt;h3&gt;
  
  
  1. "Which target European clouds have you deployed to in the last 12 months?"
&lt;/h3&gt;

&lt;p&gt;Many traditional European IT consultancies advertise "cloud migration" but do 95% of their work moving on-premises VMware servers into AWS or Azure. Ask which specific EU clouds they have automated with Terraform recently. Ask them to name two technical quirks or limitations of the target provider you are considering (for example, how private networking works across availability zones, or how load balancer proxy protocols are configured). If they cannot name specific limitations of the target cloud, they have not used it in production.&lt;/p&gt;

&lt;h3&gt;
  
  
  2. "How do you move our production database, and what is the expected write pause?"
&lt;/h3&gt;

&lt;p&gt;Listen carefully to the answer. If they propose taking a database dump (&lt;code&gt;pg_dump&lt;/code&gt; or &lt;code&gt;mysqldump&lt;/code&gt;), copying it over the network, and restoring it on the other side, your application will be offline for hours if your database is more than a few dozen gigabytes. A qualified partner will describe setting up logical replication (or native binlog/WAL replication) over an encrypted tunnel, letting the replica catch up while your application stays live, pausing writes briefly until replication lag drops near zero, and switching connection strings.&lt;/p&gt;

&lt;h3&gt;
  
  
  3. "How do you handle AWS-proprietary services like SQS, Lambda, or DynamoDB?"
&lt;/h3&gt;

&lt;p&gt;A move from EC2, RDS, and S3 to an EU cloud is mostly infrastructure plumbing. A move involving proprietary AWS services requires application code changes. Ask the partner where they draw the line between infrastructure migration and application refactoring. Do they deploy RabbitMQ, NATS, or Redis to replace SQS? Who updates the application SDK calls—their engineers or yours? Before you even speak to a vendor, you can scan your own repositories with an open-source &lt;a href="https://moveinfra.eu/tools/" rel="noopener noreferrer"&gt;cloud lock-in scanner&lt;/a&gt; to know exactly which proprietary APIs your codebase calls.&lt;/p&gt;

&lt;h3&gt;
  
  
  4. "What happens two days after cutover if we discover a critical issue?"
&lt;/h3&gt;

&lt;p&gt;Every consultancy has a plan for moving forward; few have a tested plan for moving backward. Once DNS points to the new European cloud and users start writing new rows to the new database, simply switching DNS back to AWS will result in data loss unless those new writes were replicated back to AWS. Ask how they maintain reverse replication after cutover and for how long.&lt;/p&gt;

&lt;h3&gt;
  
  
  5. "How do you help us avoid AWS data egress fees during the switch?"
&lt;/h3&gt;

&lt;p&gt;Moving terabytes of S3 objects and database snapshots out of AWS generates data transfer fees. Your partner should know how to use the AWS EU Data Act switching mechanisms or standard AWS account closure fee waivers, and how to avoid paying for multiple full-dataset copies during dry runs. You can check your own exposure and download a notice template using the free &lt;a href="https://moveinfra.eu/cloud-exit-fees/" rel="noopener noreferrer"&gt;Cloud Exit Fee Calculator&lt;/a&gt;. (Note: under Article 29 of the EU Data Act, Regulation (EU) 2023/2854, switching charges—including data egress fees for switching—are capped at cost until 12 January 2027 and banned entirely from 12 January 2027.)&lt;/p&gt;

&lt;h2&gt;
  
  
  Pricing models: how migration projects are billed
&lt;/h2&gt;

&lt;p&gt;Consultancies generally price cloud exits in one of four ways. Understanding the incentives behind each model will save you from budget overruns.&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Time and Materials (T&amp;amp;M):&lt;/strong&gt; You pay a daily or hourly rate for engineers. T&amp;amp;M is normal and reasonable when a project requires heavy application refactoring (for example, rewriting hundreds of Lambda functions or migrating off DynamoDB), because neither side can know the exact code complexity upfront. However, for standard infrastructure and database migrations, open-ended T&amp;amp;M shifts all execution risk onto you.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Two-stage (Paid Discovery, then Fixed Price):&lt;/strong&gt; You pay a small, fixed fee for a 1-to-3 week discovery phase where the consultancy audits your AWS account via read-only access and produces a migration plan, target bill estimate, and fixed price for execution. You should retain the right to walk away with the discovery deliverables if you do not like the execution quote.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Fixed Price / Milestone-based:&lt;/strong&gt; For standard workloads (VMs, Kubernetes, RDS, S3, load balancers), a mature specialist can inspect your billing export and read-only metadata, and give you a fixed price broken into milestones (e.g., 1. Inventory &amp;amp; Architecture, 2. IaC Environment Built, 3. Data Sync Live, 4. Production Cutover).&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Savings-share or Percentage of Spend:&lt;/strong&gt; Some firms charge a percentage of your annual cloud spend or a percentage of your first year's savings. Be cautious here: this model can incentivize a partner to push you toward the cheapest possible hardware setup (such as unmanaged bare metal without proper redundancy) just to inflate the theoretical savings figure. To check expected infrastructure savings objectively before entering negotiations, upload your billing export to a browser-based &lt;a href="https://moveinfra.eu/bill-analyzer/" rel="noopener noreferrer"&gt;Cloud Bill Analyzer&lt;/a&gt;.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Red flags to watch out for
&lt;/h2&gt;

&lt;p&gt;Walk away or pause negotiations if you see any of these behaviors during the sales process:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Promising "zero downtime" without inspecting your database:&lt;/strong&gt; Any migration that switches the primary database for write operations requires at least a brief pause in writes (even if it is only 15 to 60 seconds) to prevent split-brain data corruption. Anyone promising literal "zero downtime" before auditing your database schema and application connection pooling is either using marketing language or planning an unsafe cutover.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Using proprietary Terraform wrappers you cannot maintain:&lt;/strong&gt; Some consultancies deploy their own closed-source Terraform modules or custom management agents into your new cloud, then try to sell you a mandatory monthly managed-services retainer to maintain them. Insist on standard, human-readable Terraform or OpenTofu using official provider plugins.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Recommending US-owned "sovereign" regions when you asked for an EU-owned cloud:&lt;/strong&gt; If your requirement is escaping US jurisdiction (such as the US CLOUD Act) by moving to a European-headquartered provider, watch out for AWS Partners who try to steer you into the AWS European Sovereign Cloud instead. While the AWS European Sovereign Cloud keeps data and operations inside the EU, it remains owned by Amazon, a US corporation.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Asking for full Administrator IAM keys before signing an NDA and DPA:&lt;/strong&gt; For an initial assessment, a billing export or AWS &lt;code&gt;ReadOnlyAccess&lt;/code&gt; / &lt;code&gt;ViewOnlyAccess&lt;/code&gt; role is sufficient. A vendor never needs write access during scoping.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  What to require in the contract
&lt;/h2&gt;

&lt;p&gt;Once you select a partner, make sure your Statement of Work (SOW) and Master Services Agreement (MSA) cover these four areas explicitly:&lt;/p&gt;

&lt;h3&gt;
  
  
  1. Audited inventory and scope baseline
&lt;/h3&gt;

&lt;p&gt;Attach the exact inventory of AWS accounts, regions, VPCs, databases, and storage buckets to the contract. Define what happens if a new resource is discovered mid-project, and freeze non-essential architectural changes in AWS during the migration window so neither side is chasing a moving target.&lt;/p&gt;

&lt;h3&gt;
  
  
  2. Cutover criteria and rollback window
&lt;/h3&gt;

&lt;p&gt;Do not define project completion as "infrastructure deployed." Define completion as production traffic running on the destination cloud with all agreed health checks passing. Specify:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;The maximum acceptable replication lag before the cutover write pause begins (for example, under 5 seconds).&lt;/li&gt;
&lt;li&gt;The maximum planned maintenance window duration, after which an automatic abort and rollback to AWS is triggered.&lt;/li&gt;
&lt;li&gt;The post-cutover rollback period (typically 14 to 30 days) during which the old AWS environment remains on standby with reverse data synchronization active, before final decommissioning. (If your company is subject to DORA, Regulation (EU) 2022/2554, documenting these exit and fallback procedures also directly supports your regulatory ICT exit strategy requirements.)&lt;/li&gt;
&lt;/ul&gt;

&lt;h3&gt;
  
  
  3. Data handling, security, and least privilege
&lt;/h3&gt;

&lt;p&gt;Your migration partner will inevitably interact with infrastructure that holds personal or business-critical data. Require:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;A GDPR-compliant Data Processing Agreement (DPA) listing any subcontractors.&lt;/li&gt;
&lt;li&gt;All data in transit between AWS and the EU cloud to move over encrypted channels (such as WireGuard, IPsec VPN, or TLS 1.3) directly from your AWS account to your new EU cloud account—never staged on the consultancy's own servers.&lt;/li&gt;
&lt;li&gt;Checksum verification logs proving that object storage buckets were copied completely without silent corruption.&lt;/li&gt;
&lt;li&gt;A clause requiring immediate revocation of all IAM roles, SSH keys, and VPN access upon project sign-off.&lt;/li&gt;
&lt;/ul&gt;

&lt;h3&gt;
  
  
  4. Code ownership and handover deliverables
&lt;/h3&gt;

&lt;p&gt;State clearly that all Infrastructure as Code, CI/CD pipeline modifications, runbooks, and architecture diagrams become your company's intellectual property upon payment (or are licensed under an open, perpetual, royalty-free license without dependency on the vendor). Require a live handover session where your internal engineers run a &lt;code&gt;terraform plan&lt;/code&gt; and deploy an application change without the consultancy touching the keyboard.&lt;/p&gt;

&lt;h2&gt;
  
  
  Frequently asked questions
&lt;/h2&gt;

&lt;h3&gt;
  
  
  How long does a migration from AWS to a European cloud usually take?
&lt;/h3&gt;

&lt;p&gt;For a typical 10-to-300 person company running standard infrastructure (virtual machines, Kubernetes, PostgreSQL/MySQL, Redis, and S3 object storage), a well-run migration takes between 3 and 8 weeks from inventory to cutover, followed by a 30-day standby window before shutting down AWS. If the project requires rewriting code to replace proprietary services like DynamoDB or AWS Step Functions, timelines extend based on software engineering capacity.&lt;/p&gt;

&lt;h3&gt;
  
  
  Should we sign up for the destination EU cloud ourselves, or let the consultancy resell it?
&lt;/h3&gt;

&lt;p&gt;Always open the destination cloud account directly in your own company's name, with your own billing details and root credentials. Letting a consultancy resell the cloud infrastructure puts another legal and financial layer between you and your servers, complicates support escalation, and makes it harder to part ways with the consultancy later.&lt;/p&gt;

&lt;h3&gt;
  
  
  Can we claim AWS egress fee waivers if a consultancy runs our migration?
&lt;/h3&gt;

&lt;p&gt;Yes. Under AWS policy and the AWS EU Data Act Addendum, eligibility for data transfer fee reductions or waivers depends on the customer account and the nature of the workload exit, not on whether your internal staff or a third-party contractor executes the transfer. However, your company (as the account holder) must submit the formal support ticket or switching notice to AWS before moving the bulk data. Always verify the current procedure on the official AWS documentation pages before starting large transfers.&lt;/p&gt;

&lt;h3&gt;
  
  
  Do we need to rewrite our CI/CD pipelines to switch clouds?
&lt;/h3&gt;

&lt;p&gt;Usually, you only need to update the deployment target step. If you build Docker containers or application artifacts in GitHub Actions or GitLab CI, the build and test steps stay identical. The consultancy should work with your team to update the final deployment step—pointing it to the new container registry, Kubernetes cluster, or virtual servers on the European cloud.&lt;/p&gt;

&lt;h2&gt;
  
  
  Sources
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;a href="https://eur-lex.europa.eu/eli/reg/2023/2854/oj" rel="noopener noreferrer"&gt;Regulation (EU) 2023/2854 (EU Data Act) on EUR-Lex&lt;/a&gt; — Articles 25, 29, and 34 covering cloud switching contracts, notice periods, and the phase-out of switching and data egress charges.&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://eur-lex.europa.eu/eli/reg/2022/2554/oj" rel="noopener noreferrer"&gt;Regulation (EU) 2022/2554 (Digital Operational Resilience Act / DORA) on EUR-Lex&lt;/a&gt; — Requirements for ICT risk management and documented cloud exit strategies.&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://aws.amazon.com/ec2/pricing/on-demand/" rel="noopener noreferrer"&gt;AWS EC2 On-Demand &amp;amp; Data Transfer Pricing&lt;/a&gt; — Official AWS documentation detailing data transfer out rates and the process for requesting fee waivers when exiting AWS.&lt;/li&gt;
&lt;/ul&gt;




&lt;p&gt;&lt;em&gt;Drafted with AI assistance and fact-checked by the MoveInfra team. The original, kept up to date, is at &lt;a href="https://moveinfra.eu/blog/how-to-choose-a-cloud-exit-consultancy/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu&lt;/a&gt;. Free tools for pricing a move to an EU cloud: &lt;a href="https://moveinfra.eu/tools/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu/tools&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>cloud</category>
      <category>cloudmigration</category>
      <category>awsexit</category>
      <category>europeancloud</category>
    </item>
    <item>
      <title>Scanning Terraform for Lock-In Before a Cloud Migration</title>
      <dc:creator>Denys Zaruin</dc:creator>
      <pubDate>Mon, 09 Nov 2026 14:51:35 +0000</pubDate>
      <link>https://dev.to/moveinfra-eu/scanning-terraform-for-lock-in-before-a-cloud-migration-3ahf</link>
      <guid>https://dev.to/moveinfra-eu/scanning-terraform-for-lock-in-before-a-cloud-migration-3ahf</guid>
      <description>&lt;p&gt;&lt;em&gt;Originally published on &lt;a href="https://moveinfra.eu/blog/scan-terraform-for-lock-in/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;To scan your Terraform setup for cloud lock-in before a migration, extract a full list of resource types from your &lt;code&gt;.tf&lt;/code&gt; files or state file and group them into three categories: directly portable standards (like virtual machines, PostgreSQL databases, and S3-compatible buckets), managed services that need configuration changes (like message queues and serverless functions), and proprietary services that require application code rewrites (like DynamoDB, Firestore, or BigQuery). Pair this infrastructure audit with a scan of your application dependencies for provider-specific SDKs so you know exactly which engineering teams need to change code before you move to a European cloud provider.&lt;/p&gt;

&lt;h2&gt;
  
  
  Key takeaways
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Terraform does not make your infrastructure portable by default.&lt;/strong&gt; It standardizes how you write provisioning code, not the underlying cloud APIs. Moving from AWS to an EU cloud still requires rewriting provider blocks and resource definitions.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Most cloud lock-in hides in data stores and IAM, not compute.&lt;/strong&gt; Virtual machines, Kubernetes clusters, and standard relational databases move with minimal friction. Proprietary NoSQL databases and event buses require application changes.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Scan both &lt;code&gt;.tf&lt;/code&gt; files and your Terraform state.&lt;/strong&gt; Static code scans are fast and require no cloud credentials, while state file inspections catch resources generated dynamically by loops or external modules.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Check your application repositories alongside your infrastructure.&lt;/strong&gt; A resource in Terraform is only truly locked in if your application code calls its proprietary SDK directly instead of using a standard protocol like SQL, S3, or AMQP.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Automate the check in CI/CD.&lt;/strong&gt; Once you baseline your current lock-in, run a scanner in your pull requests to avoid adding new proprietary services while you plan your migration.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Why Terraform does not prevent cloud lock-in
&lt;/h2&gt;

&lt;p&gt;A common misconception among management and non-infrastructure engineers is that using Infrastructure as Code (IaC) like Terraform or OpenTofu means you can switch cloud providers by changing a single setting. In practice, Terraform uses provider-specific plugins. An &lt;code&gt;aws_instance&lt;/code&gt; on Amazon Web Services, a &lt;code&gt;google_compute_instance&lt;/code&gt; on Google Cloud, and an &lt;code&gt;upcloud_server&lt;/code&gt; on UpCloud all represent a Linux virtual machine, but they use completely different resource names, attributes, and networking models.&lt;/p&gt;

&lt;p&gt;However, rewriting Terraform HCL (HashiCorp Configuration Language) is rarely the hardest part of a cloud migration. Rewriting an &lt;code&gt;aws_db_instance&lt;/code&gt; running PostgreSQL into a managed database resource on Scaleway or OVHcloud takes an experienced engineer a few hours. &lt;/p&gt;

&lt;p&gt;Real lock-in happens when the cloud resource you provision in Terraform forces your application code to adopt a proprietary API or mental model. If your Terraform code provisions an &lt;code&gt;aws_dynamodb_table&lt;/code&gt; or a &lt;code&gt;google_bigquery_dataset&lt;/code&gt;, you cannot simply map that resource to an equivalent on Hetzner or UpCloud. To leave the hyperscaler, your developers will have to rewrite data access layers, change query languages, and migrate data schemas.&lt;/p&gt;

&lt;p&gt;Scanning your Terraform codebase before a migration gives you an honest bill of materials. It separates the infrastructure that can be rebuilt on a European cloud in days from the architectural dependencies that require weeks of engineering work.&lt;/p&gt;

&lt;h2&gt;
  
  
  The three tiers of Terraform resource portability
&lt;/h2&gt;

&lt;p&gt;When auditing your Terraform repositories, you do not need to evaluate every resource from scratch. Almost every cloud resource falls into one of three portability tiers.&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Portability Tier&lt;/th&gt;
&lt;th&gt;What It Means&lt;/th&gt;
&lt;th&gt;Common Terraform Examples (AWS / GCP / Azure)&lt;/th&gt;
&lt;th&gt;EU Cloud Migration Effort&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Tier 1: Directly Portable (Green)&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Uses open standards or protocols (Linux, SQL, Kubernetes API, S3 API, Redis protocol). Application code only needs a new connection string or endpoint URL.&lt;/td&gt;
&lt;td&gt;
&lt;code&gt;aws_instance&lt;/code&gt;, &lt;code&gt;aws_db_instance&lt;/code&gt; (Postgres/MySQL), &lt;code&gt;aws_s3_bucket&lt;/code&gt;, &lt;code&gt;aws_eks_cluster&lt;/code&gt;, &lt;code&gt;google_sql_database_instance&lt;/code&gt;, &lt;code&gt;azurerm_kubernetes_cluster&lt;/code&gt;, &lt;code&gt;aws_elasticache_cluster&lt;/code&gt; (Redis)&lt;/td&gt;
&lt;td&gt;
&lt;strong&gt;Low.&lt;/strong&gt; Rebuild the Terraform resource for the target EU provider; sync data via logical replication or S3 tools.&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Tier 2: Adapter Required (Yellow)&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Standard concept, but configured via proprietary APIs or deeply tied to provider networking and identity. Requires plumbing or CI/CD changes, and sometimes minor SDK swaps.&lt;/td&gt;
&lt;td&gt;
&lt;code&gt;aws_sqs_queue&lt;/code&gt;, &lt;code&gt;aws_lambda_function&lt;/code&gt;, &lt;code&gt;aws_alb&lt;/code&gt;, &lt;code&gt;google_cloudfunctions_function&lt;/code&gt;, &lt;code&gt;azurerm_service_bus_namespace&lt;/code&gt;, &lt;code&gt;aws_iam_role&lt;/code&gt;, &lt;code&gt;aws_kms_key&lt;/code&gt;
&lt;/td&gt;
&lt;td&gt;
&lt;strong&gt;Medium.&lt;/strong&gt; Replace with open equivalents (e.g., RabbitMQ, Kafka, Knative, or standard container deployments) and update app configuration.&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Tier 3: High Lock-In (Red)&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Proprietary database engines, global serverless platforms, or analytics warehouses with custom query dialects and SDKs. No drop-in EU cloud equivalent exists.&lt;/td&gt;
&lt;td&gt;
&lt;code&gt;aws_dynamodb_table&lt;/code&gt;, &lt;code&gt;google_firestore_database&lt;/code&gt;, &lt;code&gt;azurerm_cosmosdb_account&lt;/code&gt;, &lt;code&gt;aws_redshift_cluster&lt;/code&gt;, &lt;code&gt;google_bigquery_dataset&lt;/code&gt;, &lt;code&gt;aws_step_functions_state_machine&lt;/code&gt;
&lt;/td&gt;
&lt;td&gt;
&lt;strong&gt;High.&lt;/strong&gt; Requires refactoring application code, redesigning schemas (e.g., moving from DynamoDB to PostgreSQL), and writing custom data migration scripts.&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;You can check how specific hyperscaler services correspond to European alternatives using our free &lt;a href="https://moveinfra.eu/eu-cloud-map/" rel="noopener noreferrer"&gt;EU cloud service map&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  How to scan your Terraform codebase step by step
&lt;/h2&gt;

&lt;p&gt;You can audit your infrastructure in less than an hour without granting third-party tools access to your live cloud environment. Follow these four steps to build your portability inventory.&lt;/p&gt;

&lt;h3&gt;
  
  
  Step 1: Extract a raw list of resource types from your &lt;code&gt;.tf&lt;/code&gt; files
&lt;/h3&gt;

&lt;p&gt;Start by listing every resource type declared in your repository. If you use standard UNIX command-line tools (or ripgrep), you can extract and count every &lt;code&gt;resource&lt;/code&gt; block across your codebase:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;&lt;span class="nb"&gt;grep&lt;/span&gt; &lt;span class="nt"&gt;-rhE&lt;/span&gt; &lt;span class="s1"&gt;'^resource\s+"[^"]+"'&lt;/span&gt; &lt;span class="nb"&gt;.&lt;/span&gt; | &lt;span class="nb"&gt;awk&lt;/span&gt; &lt;span class="nt"&gt;-F&lt;/span&gt;&lt;span class="s1"&gt;'"'&lt;/span&gt; &lt;span class="s1"&gt;'{print $2}'&lt;/span&gt; | &lt;span class="nb"&gt;sort&lt;/span&gt; | &lt;span class="nb"&gt;uniq&lt;/span&gt; &lt;span class="nt"&gt;-c&lt;/span&gt; | &lt;span class="nb"&gt;sort&lt;/span&gt; &lt;span class="nt"&gt;-nr&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;This command outputs a ranked list of every resource type in your repository, such as:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;  42 aws_security_group_rule
  18 aws_iam_role_policy_attachment
  12 aws_s3_bucket
   6 aws_db_instance
   4 aws_sqs_queue
   2 aws_dynamodb_table
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Right away, you can ignore the noise. Having forty &lt;code&gt;aws_security_group_rule&lt;/code&gt; blocks is not a migration blocker; every European cloud has firewall rules, and translating IP ranges and ports is straightforward. Your eyes should go straight to the data stores and messaging layers at the bottom of the list: &lt;code&gt;aws_db_instance&lt;/code&gt;, &lt;code&gt;aws_sqs_queue&lt;/code&gt;, and &lt;code&gt;aws_dynamodb_table&lt;/code&gt;.&lt;/p&gt;

&lt;h3&gt;
  
  
  Step 2: Inspect the engines and attributes inside data resources
&lt;/h3&gt;

&lt;p&gt;Knowing the resource type is sometimes not enough. A single Terraform resource type can represent either a portable open-source service or a proprietary one, depending on its attributes.&lt;/p&gt;

&lt;p&gt;For example, if you see &lt;code&gt;aws_rds_cluster&lt;/code&gt; or &lt;code&gt;aws_db_instance&lt;/code&gt;, inspect the &lt;code&gt;engine&lt;/code&gt; attribute:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;code&gt;engine = "postgres"&lt;/code&gt; or &lt;code&gt;engine = "mysql"&lt;/code&gt; puts the resource firmly in &lt;strong&gt;Tier 1&lt;/strong&gt;. You can migrate this to managed PostgreSQL or MySQL on UpCloud, Scaleway, or OVHcloud using standard logical replication.&lt;/li&gt;
&lt;li&gt;
&lt;code&gt;engine = "aurora-postgresql"&lt;/code&gt; is still mostly &lt;strong&gt;Tier 1&lt;/strong&gt;, unless your application relies on Aurora-specific serverless scaling APIs or custom AWS extensions (like &lt;code&gt;aws_s3&lt;/code&gt; inside PostgreSQL to import data directly from buckets).&lt;/li&gt;
&lt;li&gt;
&lt;code&gt;engine = "sqlserver"&lt;/code&gt; or &lt;code&gt;engine = "oracle-ee"&lt;/code&gt; moves the resource into &lt;strong&gt;Tier 2 or 3&lt;/strong&gt;, because most European cloud providers do not offer managed Microsoft SQL Server or Oracle databases, requiring you to either self-host on virtual machines or migrate engines.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Similarly, check &lt;code&gt;aws_s3_bucket&lt;/code&gt; resources. Basic object storage and lifecycle expiration rules translate cleanly to European S3-compatible object storage. However, if you see &lt;code&gt;aws_s3_bucket_notification&lt;/code&gt; triggering an &lt;code&gt;aws_lambda_function&lt;/code&gt; on every file upload, you have an event-driven dependency that you will need to replicate in your application code or via an alternative webhook mechanism.&lt;/p&gt;

&lt;h3&gt;
  
  
  Step 3: Check what third-party modules are hiding
&lt;/h3&gt;

&lt;p&gt;If your engineering team uses public modules from the Terraform Registry (such as &lt;code&gt;terraform-aws-modules/vpc/aws&lt;/code&gt; or &lt;code&gt;terraform-aws-modules/eks/aws&lt;/code&gt;), a simple &lt;code&gt;grep&lt;/code&gt; on your &lt;code&gt;.tf&lt;/code&gt; files will only show &lt;code&gt;module&lt;/code&gt; blocks, hiding the underlying resources.&lt;/p&gt;

&lt;p&gt;To see every resource that Terraform actually manages—including those nested inside modules—run &lt;code&gt;terraform show&lt;/code&gt; against your state file (or a saved plan file) in JSON format and parse it with &lt;code&gt;jq&lt;/code&gt;:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;terraform show &lt;span class="nt"&gt;-json&lt;/span&gt; | jq &lt;span class="nt"&gt;-r&lt;/span&gt; &lt;span class="s1"&gt;'
  .. | objects | select(has("type") and has("mode")) 
  | select(.mode == "managed") 
  | .type
'&lt;/span&gt; | &lt;span class="nb"&gt;sort&lt;/span&gt; | &lt;span class="nb"&gt;uniq&lt;/span&gt; &lt;span class="nt"&gt;-c&lt;/span&gt; | &lt;span class="nb"&gt;sort&lt;/span&gt; &lt;span class="nt"&gt;-nr&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;This command queries the resolved state tree and prints the exact count of every managed resource currently running in your account. It will also reveal if a seemingly simple module is quietly provisioning proprietary KMS keys, CloudWatch log groups, or Systems Manager parameters behind the scenes.&lt;/p&gt;

&lt;h3&gt;
  
  
  Step 4: Automate the audit with an open-source lock-in scanner
&lt;/h3&gt;

&lt;p&gt;Doing this manually works well for a single repository, but most companies split their Terraform across multiple environments and mix infrastructure code with application code. &lt;/p&gt;

&lt;p&gt;To speed this up, you can use our free, open-source cloud lock-in scanner (available on our &lt;a href="https://moveinfra.eu/tools/" rel="noopener noreferrer"&gt;free tools page&lt;/a&gt; or directly on GitHub at &lt;code&gt;moveinfra/cloud-lock-in-scanner&lt;/code&gt;). It runs locally via CLI or as a GitHub Action and scans two things at once:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;Your Terraform files:&lt;/strong&gt; It identifies AWS, Google Cloud, and Azure resources and categorizes them by how difficult they are to move to standard European infrastructure.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Your application code:&lt;/strong&gt; It checks your &lt;code&gt;package.json&lt;/code&gt;, &lt;code&gt;requirements.txt&lt;/code&gt;, &lt;code&gt;go.mod&lt;/code&gt;, or &lt;code&gt;pom.xml&lt;/code&gt; files, as well as source code imports, for cloud-specific SDKs.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;Why scan both? Suppose your Terraform scan flags an &lt;code&gt;aws_s3_bucket&lt;/code&gt;. If your Node.js application accesses that bucket using the standard &lt;code&gt;@aws-sdk/client-s3&lt;/code&gt; library with a configurable &lt;code&gt;endpoint&lt;/code&gt; URL, switching to Scaleway or Hetzner Object Storage is a one-line configuration change. By contrast, if the code scanner finds &lt;code&gt;@aws-sdk/client-dynamodb&lt;/code&gt; or &lt;code&gt;@google-cloud/firestore&lt;/code&gt; imported across 35 different application files, you have an accurate measure of the refactoring work ahead.&lt;/p&gt;

&lt;h2&gt;
  
  
  A practical checklist for your Terraform lock-in audit
&lt;/h2&gt;

&lt;p&gt;When you review your scan results with your engineering team, walk through these six areas to turn the raw output into a migration plan:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;[ ] &lt;strong&gt;Databases and state:&lt;/strong&gt; List every database resource. Confirm which ones run standard open-source engines (PostgreSQL, MySQL, Redis) and flag any proprietary NoSQL or data warehouse services (DynamoDB, Firestore, Cosmos DB, BigQuery, Redshift).&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Object storage features:&lt;/strong&gt; Verify whether your buckets only use standard &lt;code&gt;GET&lt;/code&gt;/&lt;code&gt;PUT&lt;/code&gt;/&lt;code&gt;DELETE&lt;/code&gt; operations and presigned URLs, or if they rely on proprietary features like AWS S3 Object Lambda, Glacier deep archive APIs, or bucket-triggered serverless functions.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Identity and access management (IAM):&lt;/strong&gt; Count how many applications rely on cloud-native workload identity (such as AWS IAM Roles for Service Accounts in Kubernetes, or GCP Workload Identity) to authenticate to databases or queues instead of standard username/password or TLS certificate authentication.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Secrets and configuration:&lt;/strong&gt; Check for &lt;code&gt;aws_ssm_parameter&lt;/code&gt;, &lt;code&gt;aws_secretsmanager_secret&lt;/code&gt;, or &lt;code&gt;google_secret_manager_secret&lt;/code&gt;. Determine whether your applications fetch these secrets at runtime via cloud SDKs, or if they are injected as standard environment variables during deployment.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Queues and background jobs:&lt;/strong&gt; Locate all &lt;code&gt;aws_sqs_queue&lt;/code&gt;, &lt;code&gt;aws_sns_topic&lt;/code&gt;, or &lt;code&gt;google_pubsub_topic&lt;/code&gt; resources. Decide whether to replace them with a managed queue service on an EU provider, a Redis-backed job runner (like BullMQ or Sidekiq), or a self-hosted RabbitMQ/Kafka instance.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Untracked manual resources ("ClickOps"):&lt;/strong&gt; Compare your Terraform state against your monthly cloud invoice. Companies frequently create experimental databases, DNS zones, or storage buckets manually in the web console and forget to import them into Terraform. Uploading your billing export to our free &lt;a href="https://moveinfra.eu/bill-analyzer/" rel="noopener noreferrer"&gt;cloud bill analyzer&lt;/a&gt; is a reliable way to catch active services that do not appear in your git repositories.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  How to reduce lock-in before you migrate
&lt;/h2&gt;

&lt;p&gt;If your scan uncovers several Tier 2 and Tier 3 resources, you do not have to wait until migration day to fix them. In fact, decoupling your architecture while you are still on AWS, Google Cloud, or Azure makes the eventual cutover much safer.&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;Swap proprietary SDKs for standard protocols where possible.&lt;/strong&gt; If your app uses AWS SES via the proprietary AWS SDK to send transactional emails, consider switching the code to use standard SMTP credentials, or isolate the email-sending logic behind a single internal interface.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Inject secrets at deploy time, not runtime.&lt;/strong&gt; Instead of having your Python or Go application call AWS Secrets Manager on startup, have your CI/CD pipeline or Kubernetes External Secrets Operator fetch the secrets and mount them as standard environment variables. That way, the application code remains completely unaware of which cloud it runs on.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Consolidate small NoSQL tables into PostgreSQL.&lt;/strong&gt; Many teams use DynamoDB or Firestore simply to store session state, feature flags, or basic JSON documents because it was easy to set up. Modern PostgreSQL with &lt;code&gt;jsonb&lt;/code&gt; columns handles these workloads easily at the scale of most 10-to-300-person companies, and every major EU cloud offers managed PostgreSQL.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Add a lock-in check to pull requests.&lt;/strong&gt; When planning a migration for the next quarter, the last thing you want is a product team introducing a new dependency on AWS Step Functions or Google Cloud Tasks next Tuesday. Running a lock-in scanner in CI alerts tech leads when a pull request adds a new proprietary Terraform resource or cloud SDK.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;Finally, if your audit confirms that your infrastructure is mostly portable and you are ready to plan your timeline, remember that EU law is on your side regarding exit costs. Under Article 29 of the EU Data Act (Regulation (EU) 2023/2854), cloud providers may charge no more than their direct cost for switching—including data egress fees—until 12 January 2027, and switching charges are banned entirely from 12 January 2027 onward. Many hyperscalers also have specific exit waiver programs today, though you should always check the provider's official documentation and our &lt;a href="https://moveinfra.eu/cloud-exit-fees/" rel="noopener noreferrer"&gt;cloud exit fees guide&lt;/a&gt; before initiating a large data transfer.&lt;/p&gt;

&lt;h2&gt;
  
  
  Frequently asked questions
&lt;/h2&gt;

&lt;h3&gt;
  
  
  Does switching from Terraform to OpenTofu reduce cloud lock-in?
&lt;/h3&gt;

&lt;p&gt;No. Switching from Terraform to OpenTofu removes licensing lock-in with HashiCorp (IBM), because OpenTofu is open-source and community-governed. However, OpenTofu uses the exact same cloud provider plugins (&lt;code&gt;aws&lt;/code&gt;, &lt;code&gt;google&lt;/code&gt;, &lt;code&gt;azurerm&lt;/code&gt;). Your infrastructure code is still bound to the underlying cloud provider's APIs until you rewrite the resources for an alternative provider.&lt;/p&gt;

&lt;h3&gt;
  
  
  Can an automated tool convert my AWS Terraform files into Hetzner or UpCloud Terraform files?
&lt;/h3&gt;

&lt;p&gt;Large language models and conversion scripts can translate individual resource blocks—such as turning an &lt;code&gt;aws_instance&lt;/code&gt; into an &lt;code&gt;upcloud_server&lt;/code&gt;—reasonably well. However, a 1-to-1 file translation rarely works for a whole environment because networking models differ between clouds. AWS requires explicit route tables, internet gateways, and NAT gateways per availability zone, whereas most European clouds use simpler VPC and firewall primitives. It is cleaner to generate fresh, idiomatic Terraform modules designed for the target EU cloud that match your existing sizing and topology.&lt;/p&gt;

&lt;h3&gt;
  
  
  What if part of our infrastructure was created manually in the cloud console and is not in Terraform?
&lt;/h3&gt;

&lt;p&gt;A static Terraform scan will only show what is tracked in code. To catch "ClickOps" resources, inspect your cloud bill line by line or use read-only API discovery against your cloud account before planning a migration. Any production resource created by hand should be inventoried and written into your new EU cloud Terraform codebase so you eliminate configuration drift during the move.&lt;/p&gt;

&lt;h3&gt;
  
  
  Should we use a multi-cloud abstraction layer like Crossplane or Pulumi to avoid future lock-in?
&lt;/h3&gt;

&lt;p&gt;For most companies between 10 and 300 people, building a custom abstraction layer that tries to make AWS and an EU cloud look identical in code adds more maintenance overhead than it saves. You do not migrate clouds every month. A more practical approach is to rely on standard, portable building blocks—Linux VMs, Kubernetes, PostgreSQL, Redis, and S3-compatible object storage—and keep your Terraform clean and simple for whichever provider you currently use.&lt;/p&gt;

&lt;h2&gt;
  
  
  Sources
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;&lt;a href="https://developer.hashicorp.com/terraform/internals/json-format" rel="noopener noreferrer"&gt;HashiCorp Terraform Documentation — JSON Output Format (&lt;code&gt;terraform show -json&lt;/code&gt;)&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://eur-lex.europa.eu/eli/reg/2023/2854/oj" rel="noopener noreferrer"&gt;EUR-Lex — Regulation (EU) 2023/2854 (EU Data Act)&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://registry.terraform.io/providers/UpCloudLtd/upcloud/latest/docs" rel="noopener noreferrer"&gt;Terraform Registry — UpCloud Provider Documentation&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://registry.terraform.io/providers/scaleway/scaleway/latest/docs" rel="noopener noreferrer"&gt;Terraform Registry — Scaleway Provider Documentation&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://registry.terraform.io/providers/hetznercloud/hcloud/latest/docs" rel="noopener noreferrer"&gt;Terraform Registry — Hetzner Cloud Provider Documentation&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;




&lt;p&gt;&lt;em&gt;Drafted with AI assistance and fact-checked by the MoveInfra team. The original, kept up to date, is at &lt;a href="https://moveinfra.eu/blog/scan-terraform-for-lock-in/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu&lt;/a&gt;. Free tools for pricing a move to an EU cloud: &lt;a href="https://moveinfra.eu/tools/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu/tools&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>cloud</category>
      <category>terraform</category>
      <category>cloudmigration</category>
      <category>vendorlockin</category>
    </item>
    <item>
      <title>Same setup, seven clouds: server and database prices compared</title>
      <dc:creator>Denys Zaruin</dc:creator>
      <pubDate>Sun, 08 Nov 2026 12:34:20 +0000</pubDate>
      <link>https://dev.to/moveinfra-eu/same-setup-seven-clouds-server-and-database-prices-compared-lep</link>
      <guid>https://dev.to/moveinfra-eu/same-setup-seven-clouds-server-and-database-prices-compared-lep</guid>
      <description>&lt;p&gt;&lt;em&gt;Originally published on &lt;a href="https://moveinfra.eu/blog/same-setup-cloud-price-comparison/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;For a standard European production setup, public list prices on EU-owned cloud providers come out about 34% to 57% lower per month than on AWS, Microsoft Azure, or Google Cloud in our example stack, with server prices and outbound internet traffic fees driving most of the gap. As of 28 October 2026, an example web application stack with a managed PostgreSQL database costs approximately €774.30 per month on AWS, €764.76 on Azure, and €746.58 on Google Cloud, compared to €363.92 on Scaleway, €430.19 on OVHcloud, and €496.00 on UpCloud. On Hetzner, where PostgreSQL must be self-managed on a virtual server, the equivalent setup is approximately €336.85 per month. However, EU providers are not cheaper on every individual line item, and differences in hardware generations, storage IOPS, and support tiers mean that raw list prices are only the starting point for an infrastructure decision.&lt;/p&gt;

&lt;h2&gt;
  
  
  Key takeaways
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Compute and database instances cost less on EU clouds:&lt;/strong&gt; Dedicated-core virtual servers on Hetzner, OVHcloud, Scaleway, and UpCloud are consistently lower in price than comparable general-purpose instances on US hyperscalers.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Internet egress creates the widest cost gap:&lt;/strong&gt; Moving 1 TB to 10 TB of outbound data per month adds hundreds of euros to an AWS, Azure, or Google Cloud bill, while Hetzner, OVHcloud, Scaleway, and UpCloud charge €0.00 within their included traffic allowances and fair-use policies.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;EU clouds are not always the cheapest per line item:&lt;/strong&gt; UpCloud's MaxIOPS block storage (€220.00 for 1 TB) and production load balancer (€30.00) are more expensive than the base equivalents on AWS, Azure, and Google Cloud, and OVHcloud's Essential managed PostgreSQL node (€143.15) costs slightly more than AWS or Azure.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Service catalogs differ:&lt;/strong&gt; Hetzner does not offer managed PostgreSQL or a managed Kubernetes control plane, which lowers the bill only if your engineering team is willing to run those services on virtual servers themselves.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  How we compare the seven clouds
&lt;/h2&gt;

&lt;p&gt;Cloud pricing pages are notoriously hard to compare side by side. Providers bill in different currencies, use different billing intervals (by the second, hour, or month), and bundle network traffic or storage performance in different ways. &lt;/p&gt;

&lt;p&gt;The numbers in this article are taken from our open-data &lt;a href="https://moveinfra.eu/eu-cloud-prices/" rel="noopener noreferrer"&gt;EU Cloud Price Index&lt;/a&gt;. All figures are public list prices fetched on 28 October 2026, expressed in euros per month, excluding VAT, with no reserved-instance commitments, savings plans, or negotiated enterprise discounts. Where a provider lists prices in US dollars, amounts are converted at the European Central Bank reference rate of 1.1133 USD per EUR. A billing month is standardized at 730 hours, or the provider's monthly billing cap (672 hours for UpCloud). For US hyperscalers, prices reflect European regions: AWS Frankfurt (&lt;code&gt;eu-central-1&lt;/code&gt;), Azure Germany West Central, and Google Cloud Frankfurt (&lt;code&gt;europe-west3&lt;/code&gt;).&lt;/p&gt;

&lt;p&gt;Because providers update their catalogs and pricing tiers regularly, treat all numbers below as approximate as of 2026 and check each provider's official pricing page before finalizing a budget.&lt;/p&gt;

&lt;h2&gt;
  
  
  Virtual server prices across four sizes
&lt;/h2&gt;

&lt;p&gt;To compare compute costs, we look at four common virtual machine sizes ranging from 2 vCPU / 8 GB RAM up to 16 vCPU / 64 GB RAM. We selected general-purpose instances with dedicated vCPUs where specified (such as Hetzner's CCX series, OVHcloud's B3 series, Scaleway's PRO2/GP1, UpCloud's Premium plan, and the &lt;code&gt;m6a&lt;/code&gt; and &lt;code&gt;D_v4&lt;/code&gt; families on AWS and Azure).&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Server size&lt;/th&gt;
&lt;th&gt;AWS&lt;/th&gt;
&lt;th&gt;Azure&lt;/th&gt;
&lt;th&gt;Google Cloud&lt;/th&gt;
&lt;th&gt;Hetzner&lt;/th&gt;
&lt;th&gt;OVHcloud&lt;/th&gt;
&lt;th&gt;Scaleway&lt;/th&gt;
&lt;th&gt;UpCloud&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;2 vCPU, 8 GB&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€67.86 (&lt;code&gt;m6a.large&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€66.80 (&lt;code&gt;Standard_D2a_v4&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€56.61 (&lt;code&gt;e2-standard-2&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€42.99 (&lt;code&gt;CCX13&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€37.38 (&lt;code&gt;b3-8&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€40.95 (&lt;code&gt;PRO2-XXS&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€38.00 (&lt;code&gt;PREMIUM-2xCPU-8GB&lt;/code&gt;)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;4 vCPU, 16 GB&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€135.73 (&lt;code&gt;m6a.xlarge&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€133.59 (&lt;code&gt;Standard_D4a_v4&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€113.22 (&lt;code&gt;e2-standard-4&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€85.99 (&lt;code&gt;CCX23&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€74.68 (&lt;code&gt;b3-16&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€67.76 (&lt;code&gt;GP1-XS&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€92.00 (&lt;code&gt;PREMIUM-4xCPU-16GB&lt;/code&gt;)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;8 vCPU, 32 GB&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€271.46 (&lt;code&gt;m6a.2xlarge&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€267.25 (&lt;code&gt;Standard_D8a_v4&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€226.44 (&lt;code&gt;e2-standard-8&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€138.49 (&lt;code&gt;CCX33&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€149.36 (&lt;code&gt;b3-32&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€139.24 (&lt;code&gt;GP1-S&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€192.00 (&lt;code&gt;PREMIUM-8xCPU-32GB&lt;/code&gt;)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;16 vCPU, 64 GB&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€542.93 (&lt;code&gt;m6a.4xlarge&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€534.43 (&lt;code&gt;Standard_D16a_v4&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€452.89 (&lt;code&gt;e2-standard-16&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€275.99 (&lt;code&gt;CCX43&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€298.72 (&lt;code&gt;b3-64&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€279.97 (&lt;code&gt;GP1-M&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€384.00 (&lt;code&gt;PREMIUM-16xCPU-64GB&lt;/code&gt;)&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;&lt;em&gt;Approximate monthly list prices as of 28 October 2026, excluding VAT. Check provider pricing pages for current rates.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;Across all four tiers, AWS and Azure track within two euros of each other, and Google Cloud's &lt;code&gt;e2-standard&lt;/code&gt; line sits roughly 15% below them. However, note that Google's &lt;code&gt;E2&lt;/code&gt; series uses dynamic resource management (shared-core technology under the hood) rather than fixed physical cores.&lt;/p&gt;

&lt;p&gt;Among the European providers, compute prices are roughly 30% to 50% lower than on AWS and Azure. At the 8 vCPU / 32 GB tier, Hetzner (€138.49), Scaleway (€139.24), and OVHcloud (€149.36) are just over half the list price of AWS (€271.46). UpCloud's Premium tier is priced slightly higher than the other three EU clouds at larger sizes (€192.00 for 8 vCPU / 32 GB), reflecting its bundled local NVMe storage and high-availability SLA structure.&lt;/p&gt;

&lt;h2&gt;
  
  
  Managed PostgreSQL, load balancers, and Kubernetes
&lt;/h2&gt;

&lt;p&gt;Modern web infrastructure rarely stops at raw virtual machines. Most engineering teams rely on managed relational databases, load balancers, and container orchestration to reduce operational overhead.&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Managed service&lt;/th&gt;
&lt;th&gt;AWS&lt;/th&gt;
&lt;th&gt;Azure&lt;/th&gt;
&lt;th&gt;Google Cloud&lt;/th&gt;
&lt;th&gt;Hetzner&lt;/th&gt;
&lt;th&gt;OVHcloud&lt;/th&gt;
&lt;th&gt;Scaleway&lt;/th&gt;
&lt;th&gt;UpCloud&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Managed PostgreSQL (2 vCPU, 8 GB)&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€131.14 (&lt;code&gt;db.m7g.large&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€139.01 (&lt;code&gt;D2ds_v5&lt;/code&gt; General Purpose)&lt;/td&gt;
&lt;td&gt;€109.11 (&lt;code&gt;db-custom-2-8192&lt;/code&gt; Cloud SQL Enterprise)&lt;/td&gt;
&lt;td&gt;No managed offer&lt;/td&gt;
&lt;td&gt;€143.15 (&lt;code&gt;Essential B3-8&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€80.30 (&lt;code&gt;DB-PRO2-XXS&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€60.00 (&lt;code&gt;rdb.standard.2CPU-8GB&lt;/code&gt; Standard, 1 node)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Managed load balancer&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€17.70 (&lt;code&gt;Application Load Balancer&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€16.39 (&lt;code&gt;Standard Load Balancer&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€16.39 (&lt;code&gt;Cloud Load Balancing&lt;/code&gt; forwarding rule, first 5)&lt;/td&gt;
&lt;td&gt;€7.49 (&lt;code&gt;LB11&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€6.00 (&lt;code&gt;Octavia Load Balancer Small&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€16.79 (&lt;code&gt;LB-S&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€30.00 (&lt;code&gt;Managed Load Balancer Production S&lt;/code&gt;, 2 nodes)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Managed Kubernetes control plane&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€65.57 (&lt;code&gt;Amazon EKS&lt;/code&gt; standard support)&lt;/td&gt;
&lt;td&gt;€65.57 (&lt;code&gt;AKS Standard&lt;/code&gt; tier, Uptime SLA)&lt;/td&gt;
&lt;td&gt;€65.57 (&lt;code&gt;GKE&lt;/code&gt; cluster management fee)&lt;/td&gt;
&lt;td&gt;No managed offer&lt;/td&gt;
&lt;td&gt;€65.70 (&lt;code&gt;Managed Kubernetes Service Standard&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€80.30 (&lt;code&gt;Kapsule Dedicated-4&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€60.00 (&lt;code&gt;Managed Kubernetes Production&lt;/code&gt; control plane)&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;&lt;em&gt;Approximate monthly list prices as of 28 October 2026, excluding VAT. Load balancer prices on AWS, Azure, and Google Cloud show base fixed fees before per-GB or LCU data processing charges.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;This table highlights why you cannot assume an EU provider is automatically cheaper on every product:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Managed PostgreSQL:&lt;/strong&gt; UpCloud (€60.00) and Scaleway (€80.30) offer substantial savings over AWS (€131.14) and Azure (€139.01) for a single 2 vCPU / 8 GB database node. However, OVHcloud's &lt;code&gt;Essential B3-8&lt;/code&gt; plan is €143.15 per month, making its base list price slightly higher than both AWS and Azure. Meanwhile, Hetzner does not offer a managed database service at all. If you choose Hetzner, your team must deploy PostgreSQL on a virtual server, configure backups, and manage replication and patching yourselves.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Load balancers:&lt;/strong&gt; OVHcloud (€6.00) and Hetzner (€7.49) charge a low flat monthly fee. AWS (€17.70), Azure (€16.39), and Google Cloud (€16.39) appear mid-range in the table, though their final monthly cost will rise in production because they add usage-based fees (such as AWS Load Balancer Capacity Units) on top of the hourly base price. UpCloud's production load balancer is €30.00 per month because it provisions a dedicated two-node high-availability pair out of the box.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Kubernetes control planes:&lt;/strong&gt; If you require an SLA-backed control plane, list prices are almost identical across the industry. AWS, Azure, and Google Cloud all charge €65.57 per month, OVHcloud charges €65.70 for its Standard tier, UpCloud charges €60.00, and Scaleway's dedicated control plane (&lt;code&gt;Kapsule Dedicated-4&lt;/code&gt;) costs €80.30. (Note that OVHcloud and Scaleway also offer free shared control plane tiers, which are excluded here to keep the SLA comparison fair.)&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Storage and internet egress
&lt;/h2&gt;

&lt;p&gt;Storage and network data transfer are where cloud bills often diverge from initial architectural estimates.&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Storage or network item&lt;/th&gt;
&lt;th&gt;AWS&lt;/th&gt;
&lt;th&gt;Azure&lt;/th&gt;
&lt;th&gt;Google Cloud&lt;/th&gt;
&lt;th&gt;Hetzner&lt;/th&gt;
&lt;th&gt;OVHcloud&lt;/th&gt;
&lt;th&gt;Scaleway&lt;/th&gt;
&lt;th&gt;UpCloud&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Block storage, 1 TB SSD&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€85.50 (&lt;code&gt;EBS gp3&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€85.20 (&lt;code&gt;Premium SSD v2&lt;/code&gt; LRS provisioned)&lt;/td&gt;
&lt;td&gt;€107.80 (&lt;code&gt;Persistent Disk Balanced&lt;/code&gt; pd-balanced, zonal)&lt;/td&gt;
&lt;td&gt;€44.00 (&lt;code&gt;Volumes&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€86.00 (&lt;code&gt;Block Storage high-speed&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€94.90 (&lt;code&gt;Block Storage 5K&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€220.00 (&lt;code&gt;MaxIOPS&lt;/code&gt;)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Object storage, 1 TB&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€22.00 (&lt;code&gt;S3 Standard&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€17.60 (&lt;code&gt;Blob Storage&lt;/code&gt; General Block v2, Hot LRS)&lt;/td&gt;
&lt;td&gt;€20.70 (&lt;code&gt;Cloud Storage Standard&lt;/code&gt;, regional europe-west3)&lt;/td&gt;
&lt;td&gt;€6.40 (&lt;code&gt;Object Storage&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€14.00 (&lt;code&gt;Object Storage Standard&lt;/code&gt; S3 API, 3-AZ)&lt;/td&gt;
&lt;td&gt;€16.10 (&lt;code&gt;Object Storage Standard&lt;/code&gt; Multi-AZ)&lt;/td&gt;
&lt;td&gt;€20.00 (&lt;code&gt;Managed Object Storage&lt;/code&gt;)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Internet egress, 1 TB (from one server)&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€72.72&lt;/td&gt;
&lt;td&gt;€70.29&lt;/td&gt;
&lt;td&gt;€107.80&lt;/td&gt;
&lt;td&gt;€0.00&lt;/td&gt;
&lt;td&gt;€0.00&lt;/td&gt;
&lt;td&gt;€0.00&lt;/td&gt;
&lt;td&gt;€0.00&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Internet egress, 10 TB (from one server)&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€799.92&lt;/td&gt;
&lt;td&gt;€773.19&lt;/td&gt;
&lt;td&gt;€997.22&lt;/td&gt;
&lt;td&gt;€0.00&lt;/td&gt;
&lt;td&gt;€0.00&lt;/td&gt;
&lt;td&gt;€0.00&lt;/td&gt;
&lt;td&gt;€0.00&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;&lt;em&gt;Approximate monthly list prices as of 28 October 2026, excluding VAT. Egress €0.00 on EU providers means no excess fee within included traffic allowances and fair-use terms; check each provider's limits.&lt;/em&gt;&lt;/p&gt;

&lt;h3&gt;
  
  
  The storage nuances
&lt;/h3&gt;

&lt;p&gt;On block storage, AWS (&lt;code&gt;EBS gp3&lt;/code&gt; at €85.50), Azure (&lt;code&gt;Premium SSD v2&lt;/code&gt; at €85.20), OVHcloud (€86.00), and Scaleway (€94.90) sit in a tight band between €85 and €95 per terabyte. Hetzner is roughly half that price at €44.00 for 1 TB of network volumes. &lt;/p&gt;

&lt;p&gt;UpCloud is a clear outlier at €220.00 per month for 1 TB of &lt;code&gt;MaxIOPS&lt;/code&gt; block storage. UpCloud's default storage tier is built for high enterprise IOPS (up to 100,000 read IOPS), which makes it fast for heavy transactional workloads, but expensive if you simply need bulk disk space attached to a server.&lt;/p&gt;

&lt;p&gt;For S3-compatible object storage, prices range from €6.40 per TB on Hetzner and €14.00 on OVHcloud (3-AZ) up to €20.00 on UpCloud and €22.00 on AWS S3 Standard. Keep in mind that redundancy models differ: OVHcloud and Scaleway specify multi-availability-zone (Multi-AZ) durability in these tiers, whereas single-datacenter object storage carries a different risk profile.&lt;/p&gt;

&lt;h3&gt;
  
  
  The egress multiplier
&lt;/h3&gt;

&lt;p&gt;The bottom two rows of the table show the largest structural difference between US hyperscalers and European cloud providers: internet egress (data transferred out from your servers to the public internet).&lt;/p&gt;

&lt;p&gt;On AWS, Azure, and Google Cloud, outbound traffic after a small free tier is billed per gigabyte. Sending 1 TB of traffic from a single server costs €72.72 on AWS, €70.29 on Azure, and €107.80 on Google Cloud. At 10 TB per month—common for media-heavy SaaS applications, file downloads, or high-traffic APIs—egress alone costs €799.92 on AWS, €773.19 on Azure, and €997.22 on Google Cloud.&lt;/p&gt;

&lt;p&gt;On Hetzner, OVHcloud, Scaleway, and UpCloud, both the 1 TB and 10 TB scenarios cost €0.00 in excess fees. This does not mean bandwidth is infinite. Rather, these providers either include generous monthly traffic pools (for example, Hetzner includes at least 20 TB per server in European locations, and UpCloud pools terabytes of transfer based on your compute plan) or include unmetered public bandwidth subject to port speed caps and fair-use terms (as with OVHcloud and Scaleway). Always verify the exact included traffic limit and port speed on the provider's documentation if your workload pushes sustained high throughput.&lt;/p&gt;

&lt;h2&gt;
  
  
  Putting it together: the example stack
&lt;/h2&gt;

&lt;p&gt;Individual component prices only tell part of the story because a real application combines compute, databases, storage, and networking. In the &lt;a href="https://moveinfra.eu/eu-cloud-prices/" rel="noopener noreferrer"&gt;EU Cloud Price Index&lt;/a&gt;, we track a combined &lt;strong&gt;Example stack&lt;/strong&gt; representing a small production environment across all seven clouds: three servers with 4 vCPU and 16 GB, 500 GB of block storage, 1 TB of object storage, 2 TB of internet egress, a load balancer, and PostgreSQL with 2 vCPU and 8 GB.&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Provider&lt;/th&gt;
&lt;th&gt;Example stack monthly list price&lt;/th&gt;
&lt;th&gt;Database note&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;AWS&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€774.30&lt;/td&gt;
&lt;td&gt;Includes managed PostgreSQL&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Azure&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€764.76&lt;/td&gt;
&lt;td&gt;Includes managed PostgreSQL&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Google Cloud&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€746.58&lt;/td&gt;
&lt;td&gt;Includes managed PostgreSQL&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;UpCloud&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€496.00&lt;/td&gt;
&lt;td&gt;Includes managed PostgreSQL&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;OVHcloud&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€430.19&lt;/td&gt;
&lt;td&gt;Includes managed PostgreSQL&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Scaleway&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€363.92&lt;/td&gt;
&lt;td&gt;Includes managed PostgreSQL&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Hetzner&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€336.85&lt;/td&gt;
&lt;td&gt;No managed PostgreSQL: self-run on a 2 vCPU, 8 GB server&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;&lt;em&gt;Approximate monthly list prices as of 28 October 2026, excluding VAT.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;When you combine the building blocks into a single monthly bill, the three US hyperscalers cluster tightly between €746.58 and €774.30. The three European clouds with managed PostgreSQL (UpCloud, OVHcloud, and Scaleway) land between €363.92 and €496.00—even after accounting for UpCloud's higher block storage and load balancer rates or OVHcloud's higher database price, because lower compute and €0.00 egress pull the overall total down.&lt;/p&gt;

&lt;p&gt;Hetzner comes in lowest at €336.85, but it is not an apples-to-apples operational comparison. Because Hetzner lacks a managed database service, its stack figure substitutes a 2 vCPU / 8 GB virtual server running self-hosted PostgreSQL. Saving roughly €27 per month compared to Scaleway (€363.92) rarely justifies the engineering hours required to maintain database failovers, point-in-time recovery, and security patching yourself, unless your ops team already has automated tooling in place for self-hosted databases.&lt;/p&gt;

&lt;p&gt;If you want to see how your own production footprint compares rather than a benchmark stack, you can upload your current AWS, Azure, or Google Cloud billing export to our free &lt;a href="https://moveinfra.eu/bill-analyzer/" rel="noopener noreferrer"&gt;Cloud Bill Analyzer&lt;/a&gt;. It calculates the equivalent cost on UpCloud, Hetzner, Scaleway, and OVHcloud directly in your browser, and your billing file never leaves your device.&lt;/p&gt;

&lt;h2&gt;
  
  
  Why list price is only the start of the comparison
&lt;/h2&gt;

&lt;p&gt;Before you hand a spreadsheet of list prices to your CFO or engineering lead, factor in four practical differences that do not show up in a monthly rate table:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;Hardware generations and CPU performance:&lt;/strong&gt; A vCPU is not a standardized unit of performance. An AMD EPYC Genoa core, an older Intel Xeon core, and a shared/burstable core will handle database queries and web requests at different speeds. Sometimes a higher-priced instance lets you run fewer nodes.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Storage IOPS and latency:&lt;/strong&gt; As seen in the block storage comparison, 1 TB of disk can mean 3,000 baseline IOPS on one cloud, 5,000 IOPS on another, and up to 100,000 IOPS on UpCloud MaxIOPS. If your database is disk-bound, paying extra for low-latency storage is cheaper than over-provisioning database RAM.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Multi-AZ vs. single-datacenter availability:&lt;/strong&gt; AWS, Azure, and Google Cloud regions are built around at least three isolated Availability Zones (AZs) connected by low-latency fiber. Among EU clouds, OVHcloud and Scaleway offer 3-AZ setups in core regions (such as Paris), while other regions or providers operate single-datacenter zones or require you to span multiple metro areas for redundancy.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Support and account management:&lt;/strong&gt; Enterprise support plans, response-time SLAs, and dedicated technical account managers are priced separately on all seven clouds. Check the support tier pricing if your compliance framework (such as DORA or ISO 27001) mandates specific vendor response times.&lt;/li&gt;
&lt;/ol&gt;

&lt;h2&gt;
  
  
  Frequently asked questions
&lt;/h2&gt;

&lt;h3&gt;
  
  
  Are Reserved Instances or Savings Plans included in these numbers?
&lt;/h3&gt;

&lt;p&gt;No. All prices in this comparison are public on-demand monthly list prices as of 28 October 2026. AWS, Azure, and Google Cloud offer 1-year and 3-year commitments that can reduce compute and database costs by 30% to 60%, though they lock in your spend and do not discount internet egress fees. Several European providers (including OVHcloud and UpCloud) also offer committed-use discounts on top of their lower baseline prices.&lt;/p&gt;

&lt;h3&gt;
  
  
  Why is internet egress €0.00 on the four EU cloud providers?
&lt;/h3&gt;

&lt;p&gt;European cloud providers typically bundle a high volume of outbound data transfer into the base price of their servers or network services. In the index, 1 TB and 10 TB of monthly traffic show no excess fee on Hetzner, OVHcloud, Scaleway, and UpCloud, subject to each provider's included traffic and fair-use terms. If your workload pushes tens of terabytes per server or requires sustained multi-gigabit throughput, you should check each provider's specific traffic cap and overage rate (typically around €1.00 per extra TB on Hetzner, for instance).&lt;/p&gt;

&lt;h3&gt;
  
  
  Can you run production workloads on Hetzner without managed PostgreSQL?
&lt;/h3&gt;

&lt;p&gt;Yes, many European companies do. If you run on Hetzner, teams typically deploy PostgreSQL on dedicated-core CCX instances using automated high-availability tools (such as Patroni or CloudNativePG on self-managed Kubernetes) and stream encrypted write-ahead logs (WAL) to S3-compatible object storage for point-in-time recovery. If your team prefers not to manage database state and failovers, UpCloud, Scaleway, and OVHcloud provide native managed PostgreSQL.&lt;/p&gt;

&lt;h3&gt;
  
  
  Will I have to pay egress fees to move my data off AWS, Azure, or Google Cloud?
&lt;/h3&gt;

&lt;p&gt;Under the EU Data Act (Regulation (EU) 2023/2854, Article 29), cloud switching charges—including data egress fees incurred when migrating to another provider—cannot exceed the provider's direct cost until 12 January 2027, and are banned outright from 12 January 2027 onward. In addition, AWS, Azure, and Google Cloud all currently offer formal exit waiver programs when you close a service or subscription, though you must apply through their support teams before moving your data.&lt;/p&gt;

&lt;h2&gt;
  
  
  Sources
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;&lt;a href="https://aws.amazon.com/pricing/" rel="noopener noreferrer"&gt;Amazon Web Services Pricing (EC2, RDS, S3, EBS, EKS)&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://azure.microsoft.com/en-us/pricing/" rel="noopener noreferrer"&gt;Microsoft Azure Pricing Calculator and Product Pricing&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://cloud.google.com/pricing" rel="noopener noreferrer"&gt;Google Cloud Pricing (Compute Engine, Cloud SQL, Cloud Storage)&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://www.hetzner.com/cloud" rel="noopener noreferrer"&gt;Hetzner Cloud and Managed Services Pricing&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://www.ovhcloud.com/en-ie/public-cloud/prices/" rel="noopener noreferrer"&gt;OVHcloud Public Cloud Pricing&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://www.scaleway.com/en/pricing/" rel="noopener noreferrer"&gt;Scaleway Public Cloud Pricing&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://upcloud.com/pricing/" rel="noopener noreferrer"&gt;UpCloud Pricing (Cloud Servers, Managed Databases, Storage)&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://eur-lex.europa.eu/eli/reg/2023/2854/oj" rel="noopener noreferrer"&gt;Regulation (EU) 2023/2854 (EU Data Act) on EUR-Lex&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;




&lt;p&gt;&lt;em&gt;Drafted with AI assistance and fact-checked by the MoveInfra team. The original, kept up to date, is at &lt;a href="https://moveinfra.eu/blog/same-setup-cloud-price-comparison/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu&lt;/a&gt;. Free tools for pricing a move to an EU cloud: &lt;a href="https://moveinfra.eu/tools/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu/tools&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>cloud</category>
      <category>cloudpricing</category>
      <category>aws</category>
      <category>azure</category>
    </item>
    <item>
      <title>How much does it cost to leave AWS? Egress fees and waivers</title>
      <dc:creator>Denys Zaruin</dc:creator>
      <pubDate>Sat, 07 Nov 2026 12:13:40 +0000</pubDate>
      <link>https://dev.to/moveinfra-eu/how-much-does-it-cost-to-leave-aws-egress-fees-and-waivers-oik</link>
      <guid>https://dev.to/moveinfra-eu/how-much-does-it-cost-to-leave-aws-egress-fees-and-waivers-oik</guid>
      <description>&lt;p&gt;&lt;em&gt;Originally published on &lt;a href="https://moveinfra.eu/blog/how-much-does-it-cost-to-leave-aws/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;When calculating how much it costs to leave AWS, the data egress bill is often the smallest part of the budget and can usually be waived. At list prices as of 28 October 2026, transferring 1 TB of data out of one AWS server in Frankfurt to the internet costs €72.72, and 10 TB costs €799.92. However, AWS waives these data transfer charges if you move all your data off the platform or off a specific service by requesting a waiver through AWS Support before you start. Furthermore, under the EU Data Act, cloud switching charges are capped at the provider's direct cost until 12 January 2027, when they are banned entirely. The real costs of leaving AWS are engineering time, running two cloud environments in parallel during the transition, and any remaining financial commitments like Savings Plans or Reserved Instances.&lt;/p&gt;

&lt;h2&gt;
  
  
  Key takeaways
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Egress fees are rarely the blocker:&lt;/strong&gt; Moving 10 TB of data out of an AWS Frankfurt server costs €799.92 at list prices as of 28 October 2026, and you can get this credited back through the AWS exit waiver program or EU Data Act switching rules.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;You must ask before you move:&lt;/strong&gt; AWS only waives egress charges if you contact AWS Support and get approval before transferring your data.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Switching fees end in January 2027:&lt;/strong&gt; Under Article 29 of the EU Data Act, providers can only charge their direct cost for switching egress today, and must charge €0 starting 12 January 2027.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Budget for the transition, not just the network:&lt;/strong&gt; Your actual expenses will come from running your old and new environments in parallel for a few weeks, rewriting AWS-specific configurations, and paying off unused reserved capacity.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Lower monthly bills pay back the switch:&lt;/strong&gt; Moving a typical web stack from AWS (€774.30/month list price) to a European cloud like UpCloud (€496.00/month), OVHcloud (€430.19/month), Scaleway (€363.92/month), or Hetzner (€336.85/month) creates immediate monthly savings.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  What AWS charges for data egress at list prices
&lt;/h2&gt;

&lt;p&gt;Data egress (what AWS calls "Data Transfer Out" to the public internet) is metered per gigabyte. Whenever you copy database backups, container images, or object storage buckets from AWS to another cloud provider over the internet, that traffic counts as egress.&lt;/p&gt;

&lt;p&gt;To put the numbers in perspective, the table below shows monthly list prices fetched on 28 October 2026 from the &lt;a href="https://moveinfra.eu/eu-cloud-prices/" rel="noopener noreferrer"&gt;EU Cloud Price Index&lt;/a&gt;. All figures are in euros, excluding VAT, with no discounts or commitments, and with USD prices converted at the ECB rate of 1.1133. AWS prices are for the Frankfurt region (&lt;code&gt;eu-central-1&lt;/code&gt;), and Azure prices are for Germany West Central.&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Resource / Transfer&lt;/th&gt;
&lt;th&gt;AWS&lt;/th&gt;
&lt;th&gt;Azure&lt;/th&gt;
&lt;th&gt;Google Cloud&lt;/th&gt;
&lt;th&gt;UpCloud&lt;/th&gt;
&lt;th&gt;Hetzner&lt;/th&gt;
&lt;th&gt;Scaleway&lt;/th&gt;
&lt;th&gt;OVHcloud&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;
&lt;strong&gt;Internet egress, 1 TB&lt;/strong&gt; (from one server)&lt;/td&gt;
&lt;td&gt;€72.72&lt;/td&gt;
&lt;td&gt;€70.29&lt;/td&gt;
&lt;td&gt;€107.80&lt;/td&gt;
&lt;td&gt;€0.00*&lt;/td&gt;
&lt;td&gt;€0.00*&lt;/td&gt;
&lt;td&gt;€0.00*&lt;/td&gt;
&lt;td&gt;€0.00*&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;
&lt;strong&gt;Internet egress, 10 TB&lt;/strong&gt; (from one server)&lt;/td&gt;
&lt;td&gt;€799.92&lt;/td&gt;
&lt;td&gt;€773.19&lt;/td&gt;
&lt;td&gt;€997.22&lt;/td&gt;
&lt;td&gt;€0.00*&lt;/td&gt;
&lt;td&gt;€0.00*&lt;/td&gt;
&lt;td&gt;€0.00*&lt;/td&gt;
&lt;td&gt;€0.00*&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;
&lt;strong&gt;Object storage, 1 TB&lt;/strong&gt; (monthly storage)&lt;/td&gt;
&lt;td&gt;€22.00&lt;/td&gt;
&lt;td&gt;€17.60&lt;/td&gt;
&lt;td&gt;€20.70&lt;/td&gt;
&lt;td&gt;€20.00&lt;/td&gt;
&lt;td&gt;€6.40&lt;/td&gt;
&lt;td&gt;€16.10&lt;/td&gt;
&lt;td&gt;€14.00&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;
&lt;strong&gt;Block storage, 1 TB SSD&lt;/strong&gt; (monthly storage)&lt;/td&gt;
&lt;td&gt;€85.50&lt;/td&gt;
&lt;td&gt;€85.20&lt;/td&gt;
&lt;td&gt;€107.80&lt;/td&gt;
&lt;td&gt;€220.00&lt;/td&gt;
&lt;td&gt;€44.00&lt;/td&gt;
&lt;td&gt;€94.90&lt;/td&gt;
&lt;td&gt;€86.00&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;&lt;em&gt;*Note on €0.00 egress on EU providers: UpCloud, Hetzner, Scaleway, and OVHcloud either include generous monthly traffic allowances with their servers or offer unmetered public bandwidth under fair-use terms. Always check each provider's current pricing page and traffic limits for your specific server tier.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;For most small and mid-sized European companies (10 to 300 employees), total production data in PostgreSQL, MySQL, and S3 is between 500 GB and 10 TB. Even if you paid full list price without a waiver, moving 1 TB out of an AWS server costs €72.72, and 10 TB costs €799.92. That is a noticeable line item, but rarely enough to trap a business on a platform. &lt;/p&gt;

&lt;p&gt;If you want to estimate the exact egress cost for your own data volume and check which waivers apply, you can use our free &lt;a href="https://moveinfra.eu/cloud-exit-fees/" rel="noopener noreferrer"&gt;cloud exit fee calculator&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  How to waive AWS egress fees when leaving
&lt;/h2&gt;

&lt;p&gt;You do not need to pay standard egress rates when migrating away from AWS. Depending on whether you are leaving AWS completely, moving a single workload, or exercising your legal rights as a European customer, you have two formal paths to reduce or eliminate data transfer fees. &lt;/p&gt;

&lt;p&gt;Note that provider terms change, so always check the official AWS Service Terms and Data Transfer documentation before starting your migration.&lt;/p&gt;

&lt;h3&gt;
  
  
  Path 1: The standard AWS exit program
&lt;/h3&gt;

&lt;p&gt;Since March 2024, AWS has offered free data transfer out to the internet when a customer moves all of their data off AWS, or moves all data out of a specific AWS service (for example, moving all object storage out of Amazon S3 while keeping EC2 running).&lt;/p&gt;

&lt;p&gt;How the standard waiver works in practice:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;Apply before you move data:&lt;/strong&gt; You must open a ticket with AWS Customer Support before initiating the migration. AWS does not credit egress fees retroactively if you transfer the data first and ask later.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Provide migration details:&lt;/strong&gt; AWS Support will ask which accounts and services you are vacating, the approximate volume of data in gigabytes or terabytes, and your planned timeline.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Complete the move within the window:&lt;/strong&gt; Once approved, AWS gives you a set window (typically up to 60 or 180 days depending on the program terms applicable to your account) to export the data.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Delete the remaining data and receive credits:&lt;/strong&gt; After you move the data and delete it from the AWS service (or close the account), AWS issues billing credits to offset the data transfer charges incurred during the migration window.&lt;/li&gt;
&lt;/ol&gt;

&lt;h3&gt;
  
  
  Path 2: Switching under the EU Data Act
&lt;/h3&gt;

&lt;p&gt;If your company is based in the European Union, you have statutory switching rights under the EU Data Act (Regulation (EU) 2023/2854), which AWS incorporates into its contracts via the AWS EU Data Act Addendum.&lt;/p&gt;

&lt;p&gt;The EU Data Act sets strict rules on what cloud providers can charge and how long they can delay a move:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Switching fees (Article 29):&lt;/strong&gt; Until 12 January 2027, cloud providers may charge no more than their directly incurred cost for switching-related data egress. From 12 January 2027 onward, all switching charges (including data egress fees for moving to another provider) are completely banned.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Timelines (Article 25):&lt;/strong&gt; The maximum notice period a provider can require to initiate a switch is 2 months. Once the switch begins, the transitional period to complete the move is at most 30 calendar days (unless technically unfeasible), and your data must remain retrievable for at least 30 days after the transition ends.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Parallel run exception (Article 34):&lt;/strong&gt; Be aware of the distinction between a one-time switch and running two clouds permanently. If you run services in parallel across AWS and another provider with continuous data egress, the provider may still charge at cost for that ongoing transfer.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;To use the legal route on AWS, submit a formal Switching Request to AWS Customer Support under the AWS EU Data Act Addendum. You can find a ready-to-use notice template in our &lt;a href="https://moveinfra.eu/cloud-exit-fees/" rel="noopener noreferrer"&gt;cloud exit fees tool&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  The real costs of leaving AWS
&lt;/h2&gt;

&lt;p&gt;When CFOs and engineering leads review a cloud migration after it finishes, data egress is rarely what drove the cost. The actual budget goes into three areas: parallel infrastructure, contract commitments, and engineering hours.&lt;/p&gt;

&lt;h3&gt;
  
  
  1. Running two clouds in parallel
&lt;/h3&gt;

&lt;p&gt;You cannot shut down your AWS production environment on Friday and turn on a new cloud on Monday without risking data loss or extended downtime. A safe migration requires building the target environment, syncing databases via logical replication, copying object storage, testing deployments, and keeping a rollback path open.&lt;/p&gt;

&lt;p&gt;In practice, this means paying for both your AWS environment and your new European cloud environment at the same time for 2 to 6 weeks. If you keep reverse database replication running for 30 days after the DNS cutover as a safety net, you should budget for roughly one to two months of overlapping infrastructure bills.&lt;/p&gt;

&lt;h3&gt;
  
  
  2. Unused Savings Plans, Reserved Instances, and EDP commitments
&lt;/h3&gt;

&lt;p&gt;The biggest financial hurdle to leaving AWS is usually contractual, not technical. If you bought 1-year or 3-year Compute Savings Plans, EC2 Reserved Instances, or RDS Reserved Nodes, or signed an Enterprise Discount Program (EDP) with a minimum annual spend, those commitments do not disappear when you move your workloads.&lt;/p&gt;

&lt;p&gt;Even under the EU Data Act, early termination penalties for fixed-term discount contracts can still apply. Before planning your switch date, audit your AWS Billing console for:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;The exact expiration date of every active Savings Plan and Reserved Instance.&lt;/li&gt;
&lt;li&gt;Whether your reservations are "Convertible" or "Standard" (Standard EC2 Reserved Instances can sometimes be sold on the AWS Reserved Instance Marketplace, though liquidity varies).&lt;/li&gt;
&lt;li&gt;How much upfront vs. monthly commitment remains. Often, the monthly savings on an EU cloud are large enough that paying off the final two or three months of an AWS Savings Plan is still cheaper than renewing it.&lt;/li&gt;
&lt;/ul&gt;

&lt;h3&gt;
  
  
  3. Engineering time and unwinding proprietary services
&lt;/h3&gt;

&lt;p&gt;If your infrastructure relies on standard building blocks—Linux virtual machines, Docker containers on Kubernetes, PostgreSQL or MySQL, Redis caches, and S3-compatible object storage—moving off AWS is straightforward. Every major European cloud offers direct equivalents for these services.&lt;/p&gt;

&lt;p&gt;The engineering cost rises when your application code is tightly coupled to proprietary AWS services:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Databases and analytics:&lt;/strong&gt; Replacing Amazon DynamoDB, Amazon Redshift, or Aurora Serverless requires schema changes, query rewrites, and data pipeline adjustments.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Serverless and messaging:&lt;/strong&gt; Moving hundreds of AWS Lambda functions wired into API Gateway, EventBridge, and SQS takes time to repackage into standard containers or adapt to open-source message brokers like RabbitMQ or NATS.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Identity and permissions:&lt;/strong&gt; Untangling complex IAM roles, KMS encryption keys, and proprietary load balancer annotations in Terraform or application code requires careful auditing.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  How fast does an EU cloud pay back the move?
&lt;/h2&gt;

&lt;p&gt;To understand whether the transition effort is worth it, compare your ongoing monthly run rate on AWS against European providers.&lt;/p&gt;

&lt;p&gt;Below is an example production stack from our &lt;a href="https://moveinfra.eu/eu-cloud-prices/" rel="noopener noreferrer"&gt;EU Cloud Price Index&lt;/a&gt; using list prices fetched on 28 October 2026 (EUR per month, excluding VAT, no discounts, 730 hours/month or UpCloud's 672-hour cap, USD converted at 1.1133).&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Cloud Resource (Monthly List Price)&lt;/th&gt;
&lt;th&gt;AWS (Frankfurt)&lt;/th&gt;
&lt;th&gt;UpCloud&lt;/th&gt;
&lt;th&gt;OVHcloud&lt;/th&gt;
&lt;th&gt;Scaleway&lt;/th&gt;
&lt;th&gt;Hetzner&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Server: 4 vCPU, 16 GB RAM&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€135.73 (&lt;code&gt;m6a.xlarge&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€92.00&lt;/td&gt;
&lt;td&gt;€74.68 (&lt;code&gt;b3-16&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€67.76 (&lt;code&gt;GP1-XS&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€85.99 (&lt;code&gt;CCX23&lt;/code&gt;)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Server: 8 vCPU, 32 GB RAM&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€271.46 (&lt;code&gt;m6a.2xlarge&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€192.00&lt;/td&gt;
&lt;td&gt;€149.36 (&lt;code&gt;b3-32&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€139.24 (&lt;code&gt;GP1-S&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€138.49 (&lt;code&gt;CCX33&lt;/code&gt;)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Managed PostgreSQL: 2 vCPU, 8 GB&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€131.14 (&lt;code&gt;db.m7g.large&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€60.00&lt;/td&gt;
&lt;td&gt;€143.15 (&lt;code&gt;B3-8&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€80.30 (&lt;code&gt;DB-PRO2-XXS&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;&lt;em&gt;No managed offer&lt;/em&gt;&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Managed Kubernetes control plane&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€65.57 (EKS)&lt;/td&gt;
&lt;td&gt;€60.00&lt;/td&gt;
&lt;td&gt;€65.70 (MKS)&lt;/td&gt;
&lt;td&gt;€80.30 (Kapsule)&lt;/td&gt;
&lt;td&gt;&lt;em&gt;No managed offer&lt;/em&gt;&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Managed load balancer&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;€17.70 (ALB)&lt;/td&gt;
&lt;td&gt;€30.00&lt;/td&gt;
&lt;td&gt;€6.00 (Octavia S)&lt;/td&gt;
&lt;td&gt;€16.79 (&lt;code&gt;LB-S&lt;/code&gt;)&lt;/td&gt;
&lt;td&gt;€7.49 (&lt;code&gt;LB11&lt;/code&gt;)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;
&lt;strong&gt;Example complete stack&lt;/strong&gt; (with managed PostgreSQL)&lt;/td&gt;
&lt;td&gt;&lt;strong&gt;€774.30&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;&lt;strong&gt;€496.00&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;&lt;strong&gt;€430.19&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;&lt;strong&gt;€363.92&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;
&lt;strong&gt;€336.85&lt;/strong&gt;*&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;&lt;em&gt;*Note: Hetzner does not offer managed PostgreSQL or managed Kubernetes; in the Hetzner example stack (€336.85), PostgreSQL is self-hosted on a dedicated 2 vCPU, 8 GB server (€42.99/month for a &lt;code&gt;CCX13&lt;/code&gt; instance).&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;Because compute, databases, and network egress are significantly less expensive on European clouds, most companies recover the cost of running two environments in parallel within a few months of shutting down their AWS account. You can upload your own billing export to our free &lt;a href="https://moveinfra.eu/bill-analyzer/" rel="noopener noreferrer"&gt;browser-based bill analyzer&lt;/a&gt; to see what your specific resources would cost across these four providers.&lt;/p&gt;

&lt;h2&gt;
  
  
  Checklist: Before you start your AWS exit
&lt;/h2&gt;

&lt;p&gt;Use this checklist to prepare your finance and engineering teams before moving data:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;[ ] &lt;strong&gt;Audit active commitments:&lt;/strong&gt; Check AWS Billing for Savings Plans, Reserved Instances, and EDP minimum spend dates.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Measure your data footprint:&lt;/strong&gt; Total the gigabytes stored in RDS/Aurora, S3 buckets, EBS volumes, and container registries.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Map service dependencies:&lt;/strong&gt; Identify which workloads use portable tech (VMs, K8s, PostgreSQL, S3 API) and which rely on AWS-only services (DynamoDB, Lambda, SQS).&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Submit your waiver request first:&lt;/strong&gt; Open an AWS Support case requesting the data transfer out waiver—or submit a formal Switching Request under the AWS EU Data Act Addendum—&lt;em&gt;before&lt;/em&gt; syncing data.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Budget for 30–45 days of parallel run:&lt;/strong&gt; Ensure your CFO expects one to two months where both AWS and the new cloud bill at the same time.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Verify deletion before closing:&lt;/strong&gt; Once your new environment is live and your rollback period is over, delete all snapshots, S3 versions, unattached Elastic IPs, and EBS volumes so residual charges do not continue.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Frequently asked questions
&lt;/h2&gt;

&lt;h3&gt;
  
  
  Does AWS really waive egress fees when you leave?
&lt;/h3&gt;

&lt;p&gt;Yes, provided you follow their process. If you are moving all of your data off AWS or vacating a specific service completely, you must contact AWS Customer Support for approval before transferring the data. AWS provides billing credits for the approved data transfer once the move is complete and the source data is deleted. Check the AWS Data Transfer pricing page for current eligibility terms.&lt;/p&gt;

&lt;h3&gt;
  
  
  When does the EU Data Act ban cloud switching fees?
&lt;/h3&gt;

&lt;p&gt;Under Article 29 of the EU Data Act (Regulation (EU) 2023/2854), cloud switching charges—including data egress fees incurred while moving to a new provider—are banned starting 12 January 2027. Until that date, providers may only charge up to the direct cost they incur to perform the switch.&lt;/p&gt;

&lt;h3&gt;
  
  
  What happens if I just copy my S3 buckets without telling AWS Support?
&lt;/h3&gt;

&lt;p&gt;AWS will bill you standard Data Transfer Out rates automatically on your monthly invoice. At list prices as of 28 October 2026 in the Frankfurt region, transferring 1 TB from a server to the internet costs €72.72, and 10 TB costs €799.92. AWS generally does not grant retroactive waivers if you did not open a support request prior to the transfer.&lt;/p&gt;

&lt;h3&gt;
  
  
  Does moving to the AWS European Sovereign Cloud avoid lock-in and high costs?
&lt;/h3&gt;

&lt;p&gt;No. While the AWS European Sovereign Cloud keeps data and operations inside the EU, it is still owned by Amazon, a US company, and runs on proprietary AWS APIs and pricing models. If your goal is to lower your monthly infrastructure bill or remove US jurisdiction risk, moving to an EU-owned provider using standard, portable infrastructure is the more direct solution.&lt;/p&gt;

&lt;h2&gt;
  
  
  Sources
&lt;/h2&gt;

&lt;ol&gt;
&lt;li&gt;&lt;a href="https://eur-lex.europa.eu/eli/reg/2023/2854/oj" rel="noopener noreferrer"&gt;Regulation (EU) 2023/2854 (EU Data Act) — EUR-Lex&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://aws.amazon.com/ec2/pricing/on-demand/" rel="noopener noreferrer"&gt;Amazon EC2 On-Demand Pricing (Data Transfer) — AWS&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://aws.amazon.com/service-terms/" rel="noopener noreferrer"&gt;AWS Service Terms (including EU Data Act provisions) — AWS&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://aws.amazon.com/blogs/aws/free-data-transfer-out-to-internet-when-moving-out-of-aws/" rel="noopener noreferrer"&gt;Free data transfer out to internet when moving out of AWS — AWS News Blog&lt;/a&gt;&lt;/li&gt;
&lt;/ol&gt;




&lt;p&gt;&lt;em&gt;Drafted with AI assistance and fact-checked by the MoveInfra team. The original, kept up to date, is at &lt;a href="https://moveinfra.eu/blog/how-much-does-it-cost-to-leave-aws/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu&lt;/a&gt;. Free tools for pricing a move to an EU cloud: &lt;a href="https://moveinfra.eu/tools/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu/tools&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>cloud</category>
      <category>aws</category>
      <category>cloudexit</category>
      <category>egressfees</category>
    </item>
    <item>
      <title>EU-owned equivalents of AWS services: S3, RDS, EKS, Lambda and more</title>
      <dc:creator>Denys Zaruin</dc:creator>
      <pubDate>Fri, 06 Nov 2026 13:33:13 +0000</pubDate>
      <link>https://dev.to/moveinfra-eu/eu-owned-equivalents-of-aws-services-s3-rds-eks-lambda-and-more-196e</link>
      <guid>https://dev.to/moveinfra-eu/eu-owned-equivalents-of-aws-services-s3-rds-eks-lambda-and-more-196e</guid>
      <description>&lt;p&gt;&lt;em&gt;Originally published on &lt;a href="https://moveinfra.eu/blog/eu-equivalents-of-aws-services/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;European cloud providers offer direct, open-standard equivalents for the core AWS services that most applications actually run on: virtual machines (EC2), managed relational databases (RDS), S3-compatible object storage, load balancers, private networks (VPC), managed Kubernetes (EKS), and in-memory caches (ElastiCache). On UpCloud, Scaleway, and OVHcloud, you can replace these core building blocks with managed services that work with the same open protocols and Terraform workflows. Hetzner covers compute, networks, load balancers, and object storage, while databases and Kubernetes must be run on cloud servers. The real gaps appear when an architecture relies on AWS-specific services: serverless functions (Lambda) and queues (SQS/SNS) require some adjustments on most European providers, and proprietary data services like DynamoDB and Redshift have no direct managed equivalent on these four EU clouds.&lt;/p&gt;

&lt;h2&gt;
  
  
  Key takeaways
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Core infrastructure maps one-to-one:&lt;/strong&gt; EC2, S3, VPC, and Elastic Load Balancing have direct equivalents across UpCloud, Scaleway, Hetzner, and OVHcloud.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Managed databases and Kubernetes are widely available:&lt;/strong&gt; UpCloud, Scaleway, and OVHcloud offer managed PostgreSQL, MySQL, Redis/Valkey, and CNCF-certified Kubernetes. Hetzner does not offer managed databases or managed Kubernetes, so teams run them on virtual machines.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;S3 compatibility is standard:&lt;/strong&gt; All four EU providers support the S3 API for object storage, meaning existing application code and SDKs usually only need a new endpoint URL and credentials.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Serverless and messaging need a closer look:&lt;/strong&gt; Scaleway is the only provider in this group with native serverless functions and SQS/SNS-compatible queues. On the others, workloads typically move to containers or self-hosted message brokers like NATS or RabbitMQ.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Proprietary AWS databases have no drop-in replacement:&lt;/strong&gt; DynamoDB and Redshift do not have direct equivalents on these providers and require re-platforming to open-source engines like PostgreSQL or ClickHouse before or during a move.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  AWS to EU cloud service mapping
&lt;/h2&gt;

&lt;p&gt;For a complete, interactive directory that also includes Google Cloud, Azure, and DigitalOcean, see our free &lt;a href="https://moveinfra.eu/eu-cloud-map/" rel="noopener noreferrer"&gt;EU Cloud Service Map&lt;/a&gt;. The table below covers the twelve most common AWS infrastructure categories across UpCloud, Scaleway, Hetzner, and OVHcloud.&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;AWS service&lt;/th&gt;
&lt;th&gt;UpCloud&lt;/th&gt;
&lt;th&gt;Scaleway&lt;/th&gt;
&lt;th&gt;Hetzner&lt;/th&gt;
&lt;th&gt;OVHcloud&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;
&lt;strong&gt;EC2&lt;/strong&gt; (Virtual machines)&lt;/td&gt;
&lt;td&gt;Cloud Servers&lt;/td&gt;
&lt;td&gt;Instances&lt;/td&gt;
&lt;td&gt;Cloud Servers&lt;/td&gt;
&lt;td&gt;Public Cloud Instances&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;
&lt;strong&gt;RDS, Aurora (provisioned)&lt;/strong&gt; (Managed PostgreSQL / MySQL)&lt;/td&gt;
&lt;td&gt;Managed PostgreSQL / MySQL&lt;/td&gt;
&lt;td&gt;Managed Database for PostgreSQL / MySQL&lt;/td&gt;
&lt;td&gt;PostgreSQL / MySQL on Cloud Servers&lt;/td&gt;
&lt;td&gt;Managed Databases for PostgreSQL / MySQL&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;
&lt;strong&gt;S3&lt;/strong&gt; (Object storage)&lt;/td&gt;
&lt;td&gt;Managed Object Storage&lt;/td&gt;
&lt;td&gt;Object Storage&lt;/td&gt;
&lt;td&gt;Object Storage&lt;/td&gt;
&lt;td&gt;Object Storage (S3 API)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Elastic Load Balancing (ALB, NLB)&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Managed Load Balancer&lt;/td&gt;
&lt;td&gt;Load Balancer&lt;/td&gt;
&lt;td&gt;Load Balancer&lt;/td&gt;
&lt;td&gt;Public Cloud Load Balancer&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;
&lt;strong&gt;VPC, security groups&lt;/strong&gt; (Private networks, firewalls)&lt;/td&gt;
&lt;td&gt;SDN private networks, firewall rules&lt;/td&gt;
&lt;td&gt;VPC, Private Networks, security groups&lt;/td&gt;
&lt;td&gt;Networks, Firewalls&lt;/td&gt;
&lt;td&gt;vRack private networks, security groups&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;
&lt;strong&gt;EKS&lt;/strong&gt; (Kubernetes)&lt;/td&gt;
&lt;td&gt;Managed Kubernetes&lt;/td&gt;
&lt;td&gt;Kubernetes Kapsule&lt;/td&gt;
&lt;td&gt;Self-managed Kubernetes on Cloud Servers&lt;/td&gt;
&lt;td&gt;Managed Kubernetes Service&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;
&lt;strong&gt;ElastiCache&lt;/strong&gt; (Redis-compatible cache)&lt;/td&gt;
&lt;td&gt;Managed Valkey&lt;/td&gt;
&lt;td&gt;Managed Redis&lt;/td&gt;
&lt;td&gt;Valkey on Cloud Servers&lt;/td&gt;
&lt;td&gt;Managed Databases for Caching&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;
&lt;strong&gt;SQS, SNS&lt;/strong&gt; (Queues and messaging)&lt;/td&gt;
&lt;td&gt;NATS or RabbitMQ on servers&lt;/td&gt;
&lt;td&gt;Queues and Topics (SQS/SNS-compatible)&lt;/td&gt;
&lt;td&gt;NATS or RabbitMQ on servers&lt;/td&gt;
&lt;td&gt;Managed Kafka or RabbitMQ on servers&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;
&lt;strong&gt;Lambda&lt;/strong&gt; (Serverless functions)&lt;/td&gt;
&lt;td&gt;Containers on Kubernetes&lt;/td&gt;
&lt;td&gt;Serverless Functions / Containers&lt;/td&gt;
&lt;td&gt;Containers on servers&lt;/td&gt;
&lt;td&gt;Containers on Kubernetes&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;
&lt;strong&gt;Route 53&lt;/strong&gt; (DNS)&lt;/td&gt;
&lt;td&gt;EU DNS provider of your choice&lt;/td&gt;
&lt;td&gt;Domains and DNS&lt;/td&gt;
&lt;td&gt;Hetzner DNS&lt;/td&gt;
&lt;td&gt;OVHcloud DNS&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;
&lt;strong&gt;DynamoDB&lt;/strong&gt; (Proprietary NoSQL)&lt;/td&gt;
&lt;td&gt;No direct equivalent&lt;/td&gt;
&lt;td&gt;No direct equivalent&lt;/td&gt;
&lt;td&gt;No direct equivalent&lt;/td&gt;
&lt;td&gt;No direct equivalent&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;
&lt;strong&gt;Redshift&lt;/strong&gt; (Data warehouse)&lt;/td&gt;
&lt;td&gt;No direct equivalent&lt;/td&gt;
&lt;td&gt;No direct equivalent&lt;/td&gt;
&lt;td&gt;No direct equivalent&lt;/td&gt;
&lt;td&gt;No direct equivalent&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;h2&gt;
  
  
  Core infrastructure: compute, networks, and storage
&lt;/h2&gt;

&lt;p&gt;For standard web applications, SaaS backends, and APIs, the foundation of an AWS account is EC2, VPC, Elastic Load Balancing, and S3. This layer is the easiest to move because the underlying concepts are identical across clouds.&lt;/p&gt;

&lt;h3&gt;
  
  
  Virtual machines (EC2) and networking (VPC)
&lt;/h3&gt;

&lt;p&gt;AWS EC2 instances map directly to Cloud Servers on UpCloud and Hetzner, Instances on Scaleway, and Public Cloud Instances on OVHcloud. All four providers offer both shared-CPU plans for general workloads and dedicated-CPU plans for production databases and compute-heavy services, backed by local NVMe or network block storage.&lt;/p&gt;

&lt;p&gt;Networking follows the same isolation model as an AWS VPC, though with fewer moving parts. On AWS, setting up a private network requires configuring VPCs, public and private subnets, route tables, internet gateways, and often costly NAT Gateways. On UpCloud (SDN private networks), Scaleway (VPC and Private Networks), Hetzner (Networks), and OVHcloud (vRack private networks), private L2/L3 networks connect your servers across the project with far simpler routing. Stateful firewalls and security groups attach directly to servers or network interfaces to restrict traffic by IP, port, and protocol, matching the behavior of AWS security groups.&lt;/p&gt;

&lt;h3&gt;
  
  
  Load balancers (ALB and NLB)
&lt;/h3&gt;

&lt;p&gt;Elastic Load Balancing (Application Load Balancers and Network Load Balancers) maps to managed load balancers on every provider in this list. They handle TLS termination, HTTP/HTTPS routing, TCP forwarding, and health checks. If you run Kubernetes, installing an ingress controller like Envoy, NGINX, or Traefik inside the cluster automatically provisions and configures the provider's native load balancer via their Cloud Controller Manager, just as the AWS Load Balancer Controller does on EKS.&lt;/p&gt;

&lt;h3&gt;
  
  
  Object storage (S3)
&lt;/h3&gt;

&lt;p&gt;Amazon S3 is both an AWS product and the de facto industry standard API for object storage. UpCloud Managed Object Storage, Scaleway Object Storage, Hetzner Object Storage, and OVHcloud Object Storage all implement the S3 API.&lt;/p&gt;

&lt;p&gt;For most engineering teams, this means you do not have to rewrite your storage code. You continue using &lt;code&gt;boto3&lt;/code&gt; in Python, &lt;code&gt;@aws-sdk/client-s3&lt;/code&gt; in Node.js, or the AWS SDK for Go, and you change the endpoint URL and region in the client configuration. Standard features like bucket policies, pre-signed URLs, multipart uploads, and lifecycle expiration rules are supported across the board.&lt;/p&gt;

&lt;p&gt;The differences show up only if you use AWS-specific S3 extensions. If your application triggers Lambda functions directly from S3 object-creation events, uses S3 Object Lambda, or relies on AWS KMS keys tied to IAM roles for server-side encryption, you will need to move that logic into your application layer instead of relying on the storage bucket to trigger it.&lt;/p&gt;

&lt;h2&gt;
  
  
  Databases, caches, and Kubernetes
&lt;/h2&gt;

&lt;p&gt;Moving stateful workloads and container orchestration requires deciding whether you want a fully managed control plane or prefer to run the software yourself on cheaper compute.&lt;/p&gt;

&lt;h3&gt;
  
  
  Managed PostgreSQL and MySQL (RDS and Aurora)
&lt;/h3&gt;

&lt;p&gt;If you run PostgreSQL or MySQL on Amazon RDS, or if you use provisioned Amazon Aurora with PostgreSQL or MySQL compatibility, you have direct managed equivalents on UpCloud, Scaleway, and OVHcloud. These managed database services handle automated daily backups, point-in-time recovery, minor version patching, and high-availability failover across multiple nodes. Because the underlying database engine is standard open-source PostgreSQL or MySQL, data can be kept in live sync from RDS to the European cloud using native logical replication during a migration.&lt;/p&gt;

&lt;p&gt;Hetzner is the exception here. Hetzner does not offer a managed database service. Teams moving to Hetzner run PostgreSQL or MySQL directly on Cloud Servers, typically using automation tools like Ansible or Terraform, alongside open-source backup tools such as pgBackRest or Wal-G writing to object storage.&lt;/p&gt;

&lt;h3&gt;
  
  
  Redis-compatible caches (ElastiCache)
&lt;/h3&gt;

&lt;p&gt;Following Redis's licensing changes in 2024, several European providers adopted Valkey, the Linux Foundation's open-source fork of Redis that remains drop-in compatible with existing Redis clients. UpCloud offers Managed Valkey, Scaleway offers Managed Redis, and OVHcloud offers Managed Databases for Caching. Hetzner users deploy Valkey or Redis directly on Cloud Servers. Because caches normally hold ephemeral session data or temporary query results, replacing ElastiCache is usually a matter of pointing your application's connection string to the new cluster.&lt;/p&gt;

&lt;h3&gt;
  
  
  Kubernetes (EKS)
&lt;/h3&gt;

&lt;p&gt;Amazon EKS is notoriously complex because of how deeply it ties Kubernetes into AWS IAM, VPC CNI networking, and EC2 auto-scaling groups. By contrast, UpCloud Managed Kubernetes, Scaleway Kubernetes Kapsule, and OVHcloud Managed Kubernetes Service (MKS) provide standard, CNCF-certified Kubernetes clusters that feel much closer to upstream Kubernetes. You get a managed control plane, automated node pool upgrades, CSI drivers for persistent block volumes, and load balancer integration.&lt;/p&gt;

&lt;p&gt;If your workloads are already packaged as Helm charts or standard Kubernetes manifests, moving from EKS to UpCloud, Scaleway, or OVHcloud mainly involves replacing AWS-specific annotations (such as IAM Roles for Service Accounts or ALB ingress annotations) with standard Kubernetes equivalents like &lt;code&gt;cert-manager&lt;/code&gt; and a standard ingress controller. Once again, Hetzner does not offer a managed Kubernetes service, so teams using Hetzner run self-managed distributions like k3s or Talos Linux on Hetzner Cloud Servers.&lt;/p&gt;

&lt;h2&gt;
  
  
  Where you need to adapt: queues, serverless, and DNS
&lt;/h2&gt;

&lt;p&gt;Not every AWS service has a one-to-one clone on every European provider. When you move beyond VMs, databases, and Kubernetes, you occasionally have to adjust how a component is deployed.&lt;/p&gt;

&lt;h3&gt;
  
  
  Queues and messaging (SQS and SNS)
&lt;/h3&gt;

&lt;p&gt;Amazon SQS and SNS are simple, serverless message queues and pub/sub topics. If you want an exact API replacement, Scaleway offers Queues and Topics, which implements the SQS and SNS protocols directly.&lt;/p&gt;

&lt;p&gt;UpCloud, Hetzner, and OVHcloud do not have an SQS-compatible API service. On OVHcloud, you can use Managed Kafka for high-throughput event streaming, or run RabbitMQ on servers. On UpCloud and Hetzner, teams typically replace SQS and SNS by running lightweight, open-source message brokers like NATS (with JetStream) or RabbitMQ on virtual machines or inside their Kubernetes cluster, or by moving background jobs to a Redis/Valkey-backed queue such as BullMQ, Sidekiq, or Celery.&lt;/p&gt;

&lt;h3&gt;
  
  
  Serverless functions (Lambda)
&lt;/h3&gt;

&lt;p&gt;AWS Lambda encourages building applications as dozens of small, event-driven functions wired together with API Gateway and IAM. Among the four providers covered here, only Scaleway offers native Serverless Functions and Serverless Containers (built on Knative) that scale down to zero.&lt;/p&gt;

&lt;p&gt;On UpCloud, Hetzner, and OVHcloud, there is no native Function-as-a-Service product. In practice, teams migrating off Lambda package their function code into lightweight HTTP containers and run them as deployments on Managed Kubernetes (UpCloud, OVHcloud) or as containers on cloud servers (Hetzner). For scheduled cron jobs that used to run on EventBridge and Lambda, Kubernetes &lt;code&gt;CronJobs&lt;/code&gt; or systemd timers on a VM replace them cleanly.&lt;/p&gt;

&lt;h3&gt;
  
  
  DNS (Route 53)
&lt;/h3&gt;

&lt;p&gt;Standard DNS hosting is available via Scaleway Domains and DNS, Hetzner DNS, and OVHcloud DNS. UpCloud does not operate its own public authoritative DNS service, so UpCloud customers use an external European DNS provider (or keep DNS on a dedicated DNS host). If you rely on Route 53 for standard A, AAAA, CNAME, and TXT records, migrating is a straightforward zone file export and import. If you rely on Route 53 latency-based routing or health-check failover, make sure your chosen DNS provider supports traffic steering before moving the zone.&lt;/p&gt;

&lt;h2&gt;
  
  
  The hard gaps: DynamoDB and Redshift
&lt;/h2&gt;

&lt;p&gt;It is important to be upfront about where European cloud catalogs stop. Neither UpCloud, Scaleway, Hetzner, nor OVHcloud offers a direct managed equivalent to &lt;strong&gt;Amazon DynamoDB&lt;/strong&gt; or &lt;strong&gt;Amazon Redshift&lt;/strong&gt;.&lt;/p&gt;

&lt;p&gt;If your application is built heavily on DynamoDB's single-table design, DynamoDB Streams, or IAM-scoped item permissions, you cannot simply lift and shift that database to an EU cloud. Most teams with modest DynamoDB workloads (under a few hundred gigabytes and predictable query patterns) migrate that data into PostgreSQL using &lt;code&gt;jsonb&lt;/code&gt; columns and indexed keys. Teams with massive, high-throughput key-value needs either look at self-hosting ScyllaDB or Cassandra on NVMe instances, or keep that specific component separate until they can refactor the data layer.&lt;/p&gt;

&lt;p&gt;Similarly, if you run analytical workloads on Amazon Redshift, there is no drop-in managed Redshift service on these providers. Teams typically replace Redshift by deploying an open-source analytical database like ClickHouse on dedicated cloud servers, or by using PostgreSQL with columnar extensions if the dataset fits on a single large node.&lt;/p&gt;

&lt;p&gt;If you want to know exactly which AWS services your codebase and Terraform state depend on before planning a move, you can run our free, open-source cloud lock-in scanner from our &lt;a href="https://moveinfra.eu/tools/" rel="noopener noreferrer"&gt;free cloud migration tools&lt;/a&gt; page, or upload a billing export to the &lt;a href="https://moveinfra.eu/bill-analyzer/" rel="noopener noreferrer"&gt;Cloud Bill Analyzer&lt;/a&gt; to compare your current AWS spend against EU cloud list prices.&lt;/p&gt;

&lt;h2&gt;
  
  
  Checklist: auditing your AWS architecture for an EU move
&lt;/h2&gt;

&lt;p&gt;Use this five-step checklist to see how cleanly your current AWS environment maps to an EU cloud:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;List your stateful stores:&lt;/strong&gt; Check whether your data lives in RDS (PostgreSQL/MySQL), ElastiCache, and S3, or if you have dependencies on DynamoDB, Aurora Serverless, or Redshift.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Check how you use S3:&lt;/strong&gt; Verify that your app talks to S3 via standard SDK methods (&lt;code&gt;GetObject&lt;/code&gt;, &lt;code&gt;PutObject&lt;/code&gt;, pre-signed URLs) rather than triggering AWS Lambda functions directly from bucket events.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Inspect your compute layer:&lt;/strong&gt; Determine whether your apps run on EC2, ECS, or EKS (which map cleanly to VMs and Kubernetes) or if your core API is built entirely out of Lambda functions and API Gateway.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Decide between managed services and raw compute:&lt;/strong&gt; If you want managed PostgreSQL and managed Kubernetes, shortlist UpCloud, Scaleway, or OVHcloud. If your team prefers managing its own databases and Kubernetes clusters on bare VMs for lower raw compute prices, include Hetzner.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Review IAM coupled into application code:&lt;/strong&gt; Look for places where your application uses AWS STS or IAM roles to authenticate between internal services, and plan to replace them with standard service accounts, mTLS, or environment credentials.&lt;/li&gt;
&lt;/ol&gt;

&lt;h2&gt;
  
  
  Frequently asked questions
&lt;/h2&gt;

&lt;h3&gt;
  
  
  Can I keep using my existing S3 code when moving to an EU cloud?
&lt;/h3&gt;

&lt;p&gt;Yes, in most cases. UpCloud, Scaleway, Hetzner, and OVHcloud all provide S3-compatible object storage APIs. You can keep using the official AWS SDKs (such as &lt;code&gt;boto3&lt;/code&gt; or the AWS SDK for JavaScript) by changing the configured endpoint URL, region name, and access keys. You only need to change code if you rely on AWS-only features like S3 event notifications tied to Lambda or IAM-based bucket access.&lt;/p&gt;

&lt;h3&gt;
  
  
  What about the AWS European Sovereign Cloud?
&lt;/h3&gt;

&lt;p&gt;The AWS European Sovereign Cloud keeps data residency and day-to-day operations inside the European Union, which helps with physical location requirements. However, it is still owned by Amazon, a US corporation, and runs the same proprietary AWS services and pricing structure. If your goal is to avoid US legal jurisdiction (such as the US CLOUD Act), reduce cloud bills, or move onto open-standard infrastructure, an EU-owned provider is a different category of choice.&lt;/p&gt;

&lt;h3&gt;
  
  
  How do I replace AWS IAM on a European cloud?
&lt;/h3&gt;

&lt;p&gt;European clouds do not attempt to replicate the thousands of granular resource permissions in AWS IAM. Instead, they offer project-level access control and API tokens for managing infrastructure, while application-level authentication is handled using standard protocols: database usernames and passwords, dedicated S3 access keys per bucket, Kubernetes RBAC, and OpenID Connect (OIDC).&lt;/p&gt;

&lt;h3&gt;
  
  
  Do I have to pay AWS data egress fees when I move my S3 buckets and RDS databases?
&lt;/h3&gt;

&lt;p&gt;Under the EU Data Act (Regulation (EU) 2023/2854), cloud switching charges, including data egress fees when moving to another provider, cannot exceed the provider's direct cost and are banned entirely from 12 January 2027. In addition, AWS already offers a waiver for data transfer out when you move all your data off AWS or off a specific AWS service after contacting AWS Support, as well as switching terms under the AWS EU Data Act Addendum. Always check the current AWS documentation before starting a large data transfer.&lt;/p&gt;

&lt;h2&gt;
  
  
  Sources
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;&lt;a href="https://eur-lex.europa.eu/eli/reg/2023/2854/oj" rel="noopener noreferrer"&gt;Regulation (EU) 2023/2854 (EU Data Act) on EUR-Lex&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://upcloud.com/docs/" rel="noopener noreferrer"&gt;UpCloud Documentation: Products and Guides&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://www.scaleway.com/en/docs/" rel="noopener noreferrer"&gt;Scaleway Documentation&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://docs.hetzner.com/" rel="noopener noreferrer"&gt;Hetzner Documentation&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://help.ovhcloud.com/csm/en-documentation" rel="noopener noreferrer"&gt;OVHcloud Documentation&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;




&lt;p&gt;&lt;em&gt;Drafted with AI assistance and fact-checked by the MoveInfra team. The original, kept up to date, is at &lt;a href="https://moveinfra.eu/blog/eu-equivalents-of-aws-services/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu&lt;/a&gt;. Free tools for pricing a move to an EU cloud: &lt;a href="https://moveinfra.eu/tools/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu/tools&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>cloud</category>
      <category>aws</category>
      <category>eucloud</category>
      <category>cloudmigration</category>
    </item>
    <item>
      <title>EU Region vs EU-Owned Cloud: What Your Data's Address Misses</title>
      <dc:creator>Denys Zaruin</dc:creator>
      <pubDate>Thu, 05 Nov 2026 20:28:55 +0000</pubDate>
      <link>https://dev.to/moveinfra-eu/eu-region-vs-eu-owned-cloud-what-your-datas-address-misses-1dn5</link>
      <guid>https://dev.to/moveinfra-eu/eu-region-vs-eu-owned-cloud-what-your-datas-address-misses-1dn5</guid>
      <description>&lt;p&gt;&lt;em&gt;Originally published on &lt;a href="https://moveinfra.eu/blog/eu-region-vs-eu-owned-cloud/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;Hosting your infrastructure in an EU region of a US cloud provider—such as AWS &lt;code&gt;eu-central-1&lt;/code&gt; in Frankfurt or Azure West Europe in Amsterdam—means your servers sit physically inside the European Union, which gives you data residency. Using an EU-owned cloud—such as UpCloud, Scaleway, OVHcloud, or Hetzner—means both the physical servers and the parent company are headquartered in Europe and governed solely by European law, which gives you jurisdictional sovereignty. For many European companies, an EU region is enough to lowers network latency and satisfy basic data location rules. However, if your compliance requirements, customer contracts, or risk models require immunity from extraterritorial foreign laws like the US CLOUD Act, physical location alone does not solve the problem.&lt;/p&gt;

&lt;h2&gt;
  
  
  Key takeaways
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Data residency is not data sovereignty:&lt;/strong&gt; Residency is where your data is physically stored; sovereignty is which country’s laws control the company that operates the servers.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;US parent companies remain under US jurisdiction:&lt;/strong&gt; Under the US CLOUD Act, US authorities can compel US cloud providers to hand over data under their control, regardless of whether that data is stored in Frankfurt, Dublin, or Virginia.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;"Sovereign" compartments still have US roots:&lt;/strong&gt; Products like the AWS European Sovereign Cloud isolate operations and staff inside the EU, but the ultimate parent entity remains a US corporation.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Not all workloads need the same level of control:&lt;/strong&gt; Many B2B SaaS teams can run safely in an EU region of a US cloud, while teams selling to the European public sector, healthcare, or defense increasingly need an EU-owned provider.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Exit rules are tightening:&lt;/strong&gt; The EU Data Act and DORA are making it legally and financially easier to switch cloud providers, removing traditional lock-in barriers by January 2027.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Residency vs. jurisdiction: two different questions
&lt;/h2&gt;

&lt;p&gt;When a prospective enterprise customer or data protection officer asks, "Where is our data hosted?", they are usually asking two separate questions at once.&lt;/p&gt;

&lt;p&gt;The first question is &lt;strong&gt;physical&lt;/strong&gt;: &lt;em&gt;In which city and country do the hard drives holding our database sit?&lt;/em&gt; This is &lt;strong&gt;data residency&lt;/strong&gt;. If you select the Frankfurt, Stockholm, or Paris region in your AWS, Google Cloud, or Azure console, your customer data stays on servers inside that geographic boundary (unless you configure cross-region backups or global services like CDNs and IAM that replicate metadata globally). &lt;/p&gt;

&lt;p&gt;The second question is &lt;strong&gt;legal&lt;/strong&gt;: &lt;em&gt;Which government has the authority to compel the cloud provider to grant access to, freeze, or hand over our systems?&lt;/em&gt; This is &lt;strong&gt;jurisdiction&lt;/strong&gt; (often called data sovereignty). Jurisdiction follows corporate ownership and control, not just server racks.&lt;/p&gt;

&lt;p&gt;If a cloud provider is headquartered in the United States—or is a wholly owned subsidiary of a US corporation—it is subject to US law alongside the local laws of the countries where it operates. If a cloud provider is headquartered in Finland, France, or Germany, with no corporate parent outside Europe, it is subject to national and EU law.&lt;/p&gt;

&lt;h2&gt;
  
  
  How extraterritorial laws cross physical borders
&lt;/h2&gt;

&lt;p&gt;To understand why the parent company's passport matters, you have to look at how modern legal frameworks treat cloud providers.&lt;/p&gt;

&lt;h3&gt;
  
  
  The US CLOUD Act and FISA Section 702
&lt;/h3&gt;

&lt;p&gt;In 2018, the United States passed the Clarifying Lawful Overseas Use of Data (CLOUD) Act. The law amended the Stored Communications Act to state that a US service provider must preserve, backup, or disclose electronic data in its "possession, custody, or control," regardless of whether that communication or record is located within or outside of the United States.&lt;/p&gt;

&lt;p&gt;Separately, Section 702 of the US Foreign Intelligence Surveillance Act (FISA) allows US intelligence agencies to compel electronic communication service providers to assist in acquiring foreign intelligence information targeting non-US persons located outside the United States.&lt;/p&gt;

&lt;p&gt;In practice, law enforcement agencies rarely go after ordinary B2B software databases, and US cloud providers publish transparency reports showing they push back against overly broad requests. Furthermore, the EU-US Data Privacy Framework (DPF), adopted by the European Commission in July 2023, currently provides a valid legal mechanism under the GDPR for transferring personal data to participating US companies, following binding commitments from the US government to limit intelligence access to what is "necessary and proportionate."&lt;/p&gt;

&lt;h3&gt;
  
  
  Why legal tension remains
&lt;/h3&gt;

&lt;p&gt;Despite the Data Privacy Framework, legal tension remains for two reasons:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;Direct conflict with GDPR Article 48:&lt;/strong&gt; Article 48 of the GDPR states that any judgment of a court or decision of an administrative authority of a third country requiring a controller or processor to transfer or disclose personal data may only be recognized or enforceable if based on an international agreement, such as a mutual legal assistance treaty (MLAT). A direct order under the US CLOUD Act bypasses the MLAT process, putting a US provider's European subsidiary in a bind between US law and EU law.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Schrems challenges:&lt;/strong&gt; The two previous transatlantic data frameworks (Safe Harbor and Privacy Shield) were invalidated by the Court of Justice of the European Union (CJEU) in the &lt;em&gt;Schrems I&lt;/em&gt; and &lt;em&gt;Schrems II&lt;/em&gt; rulings. Many European legal teams treat the current DPF as stable for now, but build infrastructure plans assuming a third challenge could happen in the future.&lt;/li&gt;
&lt;/ol&gt;

&lt;h2&gt;
  
  
  What about "EU Sovereign Cloud" offerings from US hyperscalers?
&lt;/h2&gt;

&lt;p&gt;Recognizing that European public sector bodies and regulated industries are uneasy about US jurisdiction, AWS, Microsoft, and Google have all launched specialized "sovereign" offerings in Europe.&lt;/p&gt;

&lt;p&gt;For example, the &lt;strong&gt;AWS European Sovereign Cloud&lt;/strong&gt; creates a separate physical and logical partition in Germany. It is operated exclusively by EU residents, keeps all metadata (including billing and configuration logs) inside the EU, and uses an independent European advisory board. Microsoft offers the &lt;strong&gt;EU Data Boundary&lt;/strong&gt;, and Google offers &lt;strong&gt;Sovereign Controls&lt;/strong&gt; with local partners like T-Systems or Thales.&lt;/p&gt;

&lt;p&gt;These offerings solve specific operational risks:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;They prevent support engineers outside the EU from accidentally accessing your environment during a ticket escalation.&lt;/li&gt;
&lt;li&gt;They keep operational metadata (like resource names, tags, and IAM roles) inside the EU, which standard regions sometimes replicate back to the US.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;However, they do not change the ultimate cap table. The AWS European Sovereign Cloud is still owned by Amazon.com, Inc., a US company. If the core requirement of your customer or regulator is that the provider has no corporate nexus to a non-EU jurisdiction, a sovereign partition on a US hyperscaler does not fully close that gap.&lt;/p&gt;

&lt;h2&gt;
  
  
  How to mitigate risk if you stay in an EU region
&lt;/h2&gt;

&lt;p&gt;For many companies, moving off AWS or GCP immediately isn't practical. Developers know the tools, or the application relies on proprietary services. If you stay in an EU region of a US cloud, how much protection does encryption give you?&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Standard provider-managed encryption (KMS):&lt;/strong&gt; Encrypting data at rest using the cloud provider's native Key Management Service protects you if someone steals a physical hard drive from the data center. It does not protect against a lawful order served to the provider, because the provider controls the keys.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Bring Your Own Key (BYOK):&lt;/strong&gt; You generate the encryption keys yourself and upload them to the cloud provider's hardware security module (HSM). You can revoke the key, but while your application is running, the cloud provider's infrastructure still holds the key in memory to read and write data.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Hold Your Own Key (HYOK) / External Key Management:&lt;/strong&gt; Your keys live on a separate server outside the cloud provider (for example, in your own vault or an EU-owned key manager), and data is encrypted before it ever reaches the US cloud. This satisfies strict &lt;em&gt;Schrems II&lt;/em&gt; technical measures, though it effectively reduces the cloud provider to "dumb storage" and makes it impossible to use most managed databases or compute services that need to process cleartext data.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Side-by-side comparison: EU Region vs. EU Sovereign Partition vs. EU-Owned Cloud
&lt;/h2&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Feature / Dimension&lt;/th&gt;
&lt;th&gt;US Cloud — Standard EU Region (e.g., AWS Frankfurt)&lt;/th&gt;
&lt;th&gt;US Cloud — EU Sovereign Partition (e.g., AWS European Sovereign Cloud)&lt;/th&gt;
&lt;th&gt;EU-Owned Cloud (e.g., UpCloud, Scaleway, OVHcloud, Hetzner)&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Physical location of customer data&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Inside the EU&lt;/td&gt;
&lt;td&gt;Inside the EU&lt;/td&gt;
&lt;td&gt;Inside the EU&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Location of account &amp;amp; operational metadata&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Often replicated globally / to the US&lt;/td&gt;
&lt;td&gt;Inside the EU&lt;/td&gt;
&lt;td&gt;Inside the EU&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Support &amp;amp; operations staff&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Global (follow-the-sun model)&lt;/td&gt;
&lt;td&gt;EU residents only&lt;/td&gt;
&lt;td&gt;Predominantly EU residents&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Ultimate parent company HQ&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;United States&lt;/td&gt;
&lt;td&gt;United States&lt;/td&gt;
&lt;td&gt;European Union (Finland, France, Germany, etc.)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Subject to US CLOUD Act?&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Yes&lt;/td&gt;
&lt;td&gt;Legal grey area; parent company remains US-based&lt;/td&gt;
&lt;td&gt;No (unless the provider has significant US operations/subsidiaries)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Breadth of proprietary PaaS services&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Very high (hundreds of managed services)&lt;/td&gt;
&lt;td&gt;Moderate to high (subset of standard region services)&lt;/td&gt;
&lt;td&gt;Focused on core IaaS and open-source PaaS (VMs, K8s, Postgres, S3)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Data egress pricing structure&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;High per-GB fees after small free tier (until EU Data Act rules take full effect)&lt;/td&gt;
&lt;td&gt;High per-GB fees (similar to standard regions, plus sovereign premium)&lt;/td&gt;
&lt;td&gt;Typically bundled generous transfer allowances or low flat rates (check provider pricing pages)&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;&lt;em&gt;Note on EU-owned clouds and US presence: Some larger European providers have US data centers or US sales subsidiaries. Strictly speaking, a US court can serve an order on a European company that has sufficient "minimum contacts" or a subsidiary in the US, though doing so to extract European customer data from European servers would face immediate blocking statutes and GDPR enforcement in Europe.&lt;/em&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  Beyond privacy: operational autonomy, DORA, and the EU Data Act
&lt;/h2&gt;

&lt;p&gt;For years, the conversation around EU clouds focused entirely on the GDPR and personal data. Since 2025, the focus for European CEOs and CFOs has expanded to &lt;strong&gt;operational continuity&lt;/strong&gt; and &lt;strong&gt;vendor lock-in&lt;/strong&gt;.&lt;/p&gt;

&lt;h3&gt;
  
  
  DORA (Digital Operational Resilience Act)
&lt;/h3&gt;

&lt;p&gt;Applicable since 17 January 2025, DORA (Regulation (EU) 2022/2554) governs financial entities in the EU and their critical ICT third-party service providers. If your company is a fintech, an insurtech, or a SaaS vendor selling into banks, DORA impacts you. It requires explicit exit strategies, tested migration plans, and avoidance of concentration risk for ICT services supporting critical or important functions. Relying entirely on proprietary services inside a single US hyperscaler makes passing a DORA vendor audit harder.&lt;/p&gt;

&lt;h3&gt;
  
  
  The EU Data Act and the end of egress lock-in
&lt;/h3&gt;

&lt;p&gt;Historically, moving data out of a US cloud's EU region was punishingly expensive due to data egress fees. The EU Data Act (Regulation (EU) 2023/2854) changes the economics of switching cloud providers. Under Article 29 of the Data Act:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;During the transition period until &lt;strong&gt;12 January 2027&lt;/strong&gt;, cloud providers may only pass on the direct costs incurred for switching (including data egress for switching purposes).&lt;/li&gt;
&lt;li&gt;Starting &lt;strong&gt;12 January 2027&lt;/strong&gt;, switching charges—including data egress fees when moving to another provider—are completely banned across the EU.&lt;/li&gt;
&lt;/ul&gt;

&lt;h3&gt;
  
  
  Predictable billing and open standards
&lt;/h3&gt;

&lt;p&gt;While US clouds offer hundreds of specialized services (like DynamoDB, BigQuery, or Cognito), most 10-to-300-person European technology companies actually run standard workloads: Linux virtual machines, Docker containers on Kubernetes, PostgreSQL or MySQL databases, Redis caches, and S3-compatible object storage.&lt;/p&gt;

&lt;p&gt;EU-owned providers—such as UpCloud (Finland), Scaleway and OVHcloud (France), Hetzner, STACKIT, and IONOS (Germany), and Exoscale (Switzerland/Austria)—build their platforms around these open standards. Because they do not wrap standard open-source software in complex proprietary billing meters, monthly infrastructure costs are typically lower and much easier to forecast.&lt;/p&gt;

&lt;h2&gt;
  
  
  Practical checklist: Which setup fits your company?
&lt;/h2&gt;

&lt;p&gt;Use this checklist with your engineering and legal leads to decide whether your current cloud setup matches your business reality.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Staying in a standard EU region of a US cloud makes sense if:&lt;/strong&gt;&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;[ ] Your customers are mostly private-sector businesses that accept standard GDPR Data Processing Agreements (DPAs) and the EU-US Data Privacy Framework.&lt;/li&gt;
&lt;li&gt;[ ] Your engineering team relies heavily on proprietary cloud services (e.g., AWS DynamoDB, Google BigQuery, Azure Cosmos DB) that would require rewriting application code to leave.&lt;/li&gt;
&lt;li&gt;[ ] Your primary motivation is low network latency for European users rather than geopolitical or jurisdictional insulation.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;&lt;strong&gt;Moving to an EU-owned cloud makes sense if:&lt;/strong&gt;&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;[ ] You sell (or plan to sell) to European public sector bodies, healthcare providers, defense contractors, or strict financial institutions that mandate EU-headquartered sub-processors.&lt;/li&gt;
&lt;li&gt;[ ] Your architecture runs primarily on portable building blocks: VMs, Kubernetes, PostgreSQL/MySQL, Redis, and S3-compatible storage.&lt;/li&gt;
&lt;li&gt;[ ] You need a verifiable exit strategy under DORA, or you want to eliminate single-vendor concentration risk.&lt;/li&gt;
&lt;li&gt;[ ] You want predictable monthly infrastructure bills without paying a premium for services you don't use or unexpected data transfer charges.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Frequently asked questions
&lt;/h2&gt;

&lt;h3&gt;
  
  
  Does the GDPR require me to use an EU-owned cloud?
&lt;/h3&gt;

&lt;p&gt;No. The GDPR does not ban the use of US-owned cloud providers, nor does it strictly require data to stay inside the EU, provided a valid transfer mechanism (like the EU-US Data Privacy Framework or Standard Contractual Clauses with supplementary measures) is in place. However, using an EU-owned cloud eliminates third-country transfer assessments and protects against the risk of future transatlantic data frameworks being struck down in court.&lt;/p&gt;

&lt;h3&gt;
  
  
  Can the US government really demand my data if it's stored in AWS Frankfurt?
&lt;/h3&gt;

&lt;p&gt;Legally, the US CLOUD Act gives US authorities the power to compel a US company to produce data under its control, regardless of where the servers are located. In practice, US cloud providers challenge requests that conflict with foreign laws, and the US Department of Justice generally seeks data directly from the enterprise customer first rather than the cloud provider. Still, for regulated industries, the existence of that legal authority is a compliance risk they prefer to avoid altogether.&lt;/p&gt;

&lt;h3&gt;
  
  
  Are EU-owned clouds ready for production SaaS workloads?
&lt;/h3&gt;

&lt;p&gt;Yes, for standard architectures. Providers like UpCloud, Scaleway, OVHcloud, STACKIT, IONOS, and Exoscale offer managed Kubernetes, managed PostgreSQL and MySQL with high availability, load balancers, private networking, and S3-compatible object storage, along with ISO 27001 certifications. (Hetzner offers high-performance dedicated and cloud servers at very low prices, though managed databases and Kubernetes must be set up on servers directly.) Where EU clouds differ from US hyperscalers is that they do not offer dozens of niche, proprietary AI, IoT, or global NoSQL products.&lt;/p&gt;

&lt;h3&gt;
  
  
  Will moving from an EU region to an EU-owned cloud require rewriting our application?
&lt;/h3&gt;

&lt;p&gt;If your application uses standard technologies—like Docker containers, Linux VMs, PostgreSQL, MySQL, Redis, and S3 APIs—you do not need to rewrite application code. You only need to rebuild the underlying infrastructure (networks, firewalls, compute, and managed databases) on the new provider and update your deployment pipelines. If your application is tightly coupled to proprietary services like AWS Lambda, Step Functions, and DynamoDB, migrating requires refactoring those components first.&lt;/p&gt;

&lt;h2&gt;
  
  
  Sources
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;&lt;a href="https://eur-lex.europa.eu/eli/reg/2016/679/oj" rel="noopener noreferrer"&gt;Regulation (EU) 2016/679 (General Data Protection Regulation) — EUR-Lex&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://eur-lex.europa.eu/eli/reg/2023/2854/oj" rel="noopener noreferrer"&gt;Regulation (EU) 2023/2854 (EU Data Act) — EUR-Lex&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://eur-lex.europa.eu/eli/reg/2022/2554/oj" rel="noopener noreferrer"&gt;Regulation (EU) 2022/2554 (Digital Operational Resilience Act / DORA) — EUR-Lex&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://www.dataprivacyframework.gov/" rel="noopener noreferrer"&gt;Data Privacy Framework Program — Official US Government Website&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://aws.amazon.com/compliance/europe-digital-sovereignty/" rel="noopener noreferrer"&gt;AWS European Sovereign Cloud — Official AWS Documentation&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;




&lt;p&gt;&lt;em&gt;Drafted with AI assistance and fact-checked by the MoveInfra team. The original, kept up to date, is at &lt;a href="https://moveinfra.eu/blog/eu-region-vs-eu-owned-cloud/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu&lt;/a&gt;. Free tools for pricing a move to an EU cloud: &lt;a href="https://moveinfra.eu/tools/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu/tools&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>cloud</category>
      <category>datasovereignty</category>
      <category>eucloud</category>
      <category>compliance</category>
    </item>
    <item>
      <title>Why every cloud migration needs a rollback plan, and how to build one</title>
      <dc:creator>Denys Zaruin</dc:creator>
      <pubDate>Thu, 05 Nov 2026 13:49:26 +0000</pubDate>
      <link>https://dev.to/moveinfra-eu/why-every-cloud-migration-needs-a-rollback-plan-and-how-to-build-one-l5</link>
      <guid>https://dev.to/moveinfra-eu/why-every-cloud-migration-needs-a-rollback-plan-and-how-to-build-one-l5</guid>
      <description>&lt;p&gt;&lt;em&gt;Originally published on &lt;a href="https://moveinfra.eu/blog/rollback-plan-cloud-migration/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;A cloud migration rollback plan is a documented, pre-tested procedure that lets your engineering team revert traffic and data back to your original cloud provider if serious problems appear during or after cutover. Every migration needs one because production traffic often exposes hidden issues—like network latency spikes, subtle database incompatibilities, or misconfigured firewall rules—that staging tests miss. To build a reliable rollback plan, you must keep your original environment running, lower your DNS time-to-live (TTL) before switching, pause database writes briefly during the cutover, and set up reverse data replication from the new cloud back to the old one so you can switch back days later without losing new customer data.&lt;/p&gt;

&lt;h2&gt;
  
  
  Key takeaways
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Never delete the old environment immediately:&lt;/strong&gt; Keep your source infrastructure on AWS, Google Cloud, Azure, or DigitalOcean running in a warm standby state for at least 14 to 30 days after cutover.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;State is the hardest part of a rollback:&lt;/strong&gt; Reverting stateless application servers only takes a DNS change, but reverting databases after users have written new data requires reverse logical replication.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Define objective rollback triggers in advance:&lt;/strong&gt; Agree on clear metrics—such as error rates, database CPU saturation, or payment failures—before the maintenance window begins so tired engineers do not have to debate during an outage.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Test the return path before going live:&lt;/strong&gt; A rollback plan that exists only in a document is a guess; run a dress rehearsal in staging to verify that reverse replication and DNS changes actually work.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Regulations increasingly require exit and reversal capability:&lt;/strong&gt; For financial entities and their critical ICT providers in the EU, the Digital Operational Resilience Act (DORA) requires tested exit strategies and continuity plans when switching providers.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Why cloud migrations fail after a successful cutover
&lt;/h2&gt;

&lt;p&gt;Most engineering teams spend 95% of their migration energy getting the new environment running and 5% thinking about what happens if they need to go back. When the new environment passes functional tests in staging, it is tempting to assume the cutover will be final.&lt;/p&gt;

&lt;p&gt;In practice, production environments behave differently under real user load. You might switch from a US hyperscaler to an European provider like UpCloud, Scaleway, Hetzner, or OVHcloud and find that everything works for the first three hours. Then a scheduled batch job runs, or Monday morning traffic hits, and an unexpected bottleneck appears. &lt;/p&gt;

&lt;p&gt;Common post-cutover failures include:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Connection pool exhaustion:&lt;/strong&gt; Different managed database configurations or slightly different network round-trip times between application nodes and database nodes cause connection pools to max out under peak load.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Third-party IP allowlists:&lt;/strong&gt; An external partner, banking API, or enterprise customer firewall silently blocks requests coming from your new cloud's egress IP ranges, even though your team updated the documented integrations.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Storage IOPS limits:&lt;/strong&gt; Disk read/write performance on the new virtual machines or block storage volumes hits a ceiling during heavy background processing, slowing down the entire application.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;TLS and DNS caching quirks:&lt;/strong&gt; Legacy client applications or corporate proxies ignore standard DNS TTLs and continue sending half of your customer traffic to the old environment while the other half goes to the new one.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;If you discover one of these problems ten minutes after switching DNS, fixing it forward might take hours of stressful debugging while your application is down. If you discover it three days after switching, you face a harder problem: your customers have spent three days creating orders, updating profiles, and uploading files in the new cloud. Simply pointing DNS back to the old cloud would erase those three days of business activity.&lt;/p&gt;

&lt;p&gt;A complete rollback plan solves both scenarios. It gives you an immediate escape hatch during the cutover window and a safe, data-preserving return path during the weeks that follow.&lt;/p&gt;

&lt;h2&gt;
  
  
  The two phases of a migration rollback
&lt;/h2&gt;

&lt;p&gt;To design a rollback plan that works in the real world, you need to separate the migration timeline into two distinct risk phases. Each phase has different technical requirements.&lt;/p&gt;

&lt;h3&gt;
  
  
  Phase 1: Aborting during the cutover window (Point-in-time rollback)
&lt;/h3&gt;

&lt;p&gt;The first phase covers the maintenance or cutover window itself, typically lasting from a few minutes to an hour. During this window, you stop writes to the old database, wait for replication to catch up, run sanity checks on the new environment, and switch your DNS or load balancer traffic.&lt;/p&gt;

&lt;p&gt;If something breaks before you allow real users to write data to the new database, rolling back is straightforward. Because the old database still holds the latest authoritative state, you only need to re-enable writes on the old database and keep your DNS pointing at the old cloud. No data is lost, and the old system resumes where it left off.&lt;/p&gt;

&lt;h3&gt;
  
  
  Phase 2: Reverting after live traffic hits (Stateful rollback)
&lt;/h3&gt;

&lt;p&gt;The second phase starts the exact second a user or background job writes new data to the target cloud. From this moment on, your new EU cloud is the system of record, and your old cloud is out of date.&lt;/p&gt;

&lt;p&gt;If a critical flaw appears 48 hours later—for example, a weekly reporting query crashes the database or an undocumented compliance requirement surfaces—you cannot just flip DNS back. To roll back safely during Phase 2, your old database and object storage must continuously receive every new write that happens in the new cloud. &lt;/p&gt;

&lt;h2&gt;
  
  
  How to build a stateful rollback plan step by step
&lt;/h2&gt;

&lt;p&gt;Building a reliable rollback architecture requires preparing your networks, applications, databases, and DNS well before cutover day. Here is how to set up each layer.&lt;/p&gt;

&lt;h3&gt;
  
  
  1. Lower DNS TTLs at least 48 hours early
&lt;/h3&gt;

&lt;p&gt;DNS records have a Time-To-Live (TTL) setting that tells recursive resolvers and user devices how long to cache an IP address before asking for a fresh one. If your production domain has a TTL of 3600 seconds (one hour) or 86400 seconds (24 hours), changing your DNS record during an emergency will not move traffic immediately. Many users will stay stuck on the broken environment until their local cache expires.&lt;/p&gt;

&lt;p&gt;Two to three days before your migration, lower the TTL on all relevant &lt;code&gt;A&lt;/code&gt;, &lt;code&gt;AAAA&lt;/code&gt;, and &lt;code&gt;CNAME&lt;/code&gt; records to 60 seconds (or 300 seconds if your DNS provider enforces a higher minimum). If you use a proxy service like Cloudflare or an external global load balancer, traffic shifts happen inside the proxy layer within seconds, which gives you much tighter control over both cutover and rollback.&lt;/p&gt;

&lt;h3&gt;
  
  
  2. Keep your CI/CD pipeline ready for dual deployment
&lt;/h3&gt;

&lt;p&gt;During the transition and rollback window, your application code should be identical in both clouds. Do not combine a cloud infrastructure migration with a major application rewrite or database version upgrade. &lt;/p&gt;

&lt;p&gt;Configure your CI/CD pipeline (such as GitHub Actions or GitLab CI) so that you can deploy the exact same container image or build artifact to both the old cloud and the new cloud. If you have to push an emergency bug fix two days after cutover, deploy it to the old environment too so your standby infrastructure stays ready to receive traffic.&lt;/p&gt;

&lt;h3&gt;
  
  
  3. Configure reverse database replication
&lt;/h3&gt;

&lt;p&gt;This is the technical core of a Phase 2 rollback. During the initial migration, you likely set up logical replication from your old cloud (source) to your new cloud (target) so data stays in sync up to the cutover moment.&lt;/p&gt;

&lt;p&gt;To enable a safe rollback after cutover, you must prepare the reverse path: logical replication from the new database back to the old database.&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;In PostgreSQL:&lt;/strong&gt; You use native logical replication (&lt;code&gt;CREATE PUBLICATION&lt;/code&gt; on the new database and &lt;code&gt;CREATE subscription&lt;/code&gt; on the old database). Before you switch traffic to the new cloud, you promote the new database to primary, pause writes, record the Log Sequence Number (LSN), and reverse the replication direction so the old database becomes the subscriber.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;In MySQL:&lt;/strong&gt; You configure binary log (binlog) replication with Global Transaction Identifiers (GTIDs) enabled on both sides. When the new database takes over as primary, you point the old database's replication channel (&lt;code&gt;CHANGE REPLICATION SOURCE TO&lt;/code&gt;) at the new primary so it pulls every new transaction back to the old cloud.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Verify that primary keys and auto-increment sequences do not collide, and ensure your network firewall rules allow encrypted database traffic to flow in both directions between the two clouds.&lt;/p&gt;

&lt;h3&gt;
  
  
  4. Sync new object storage writes back to the old bucket
&lt;/h3&gt;

&lt;p&gt;If your application stores user uploads, invoices, or media files in object storage (like AWS S3, Google Cloud Storage, or Azure Blob Storage) and you are moving to an S3-compatible bucket on an EU provider, you need a plan for new files created after cutover.&lt;/p&gt;

&lt;p&gt;You have two practical options:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;Continuous reverse sync:&lt;/strong&gt; Run a background worker or scheduled job (using tools like &lt;code&gt;rclone&lt;/code&gt; with checksum verification) every few minutes to copy newly created or modified objects from the new EU bucket back to the old cloud bucket.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Application dual-writing:&lt;/strong&gt; Update your storage adapter code to write new files to both buckets simultaneously during the 30-day transition period, while reading from the primary cloud.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;Avoid syncing deletions automatically to the old bucket during the rollback window. If a bug in the new environment accidentally deletes objects, you want those files to remain intact in the old bucket.&lt;/p&gt;

&lt;h3&gt;
  
  
  5. Enforce a clean write pause during the switch
&lt;/h3&gt;

&lt;p&gt;Whether you are migrating forward or rolling back, never switch active database primaries while both sides are accepting writes. Doing so causes split-brain data conflicts that take days to untangle manually.&lt;/p&gt;

&lt;p&gt;Build a simple "maintenance mode" or read-only toggle into your application or load balancer. When executing the cutover—or when executing a rollback—follow this exact sequence:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;Enable read-only / maintenance mode to stop incoming writes.&lt;/li&gt;
&lt;li&gt;Terminate any remaining active write connections on the current database.&lt;/li&gt;
&lt;li&gt;Wait until replication lag between the two clouds reaches zero (typically under 5 seconds on a healthy link).&lt;/li&gt;
&lt;li&gt;Promote the destination database and reverse the replication stream.&lt;/li&gt;
&lt;li&gt;Switch DNS or load balancer targets and disable maintenance mode.&lt;/li&gt;
&lt;/ol&gt;

&lt;h2&gt;
  
  
  Defining your rollback triggers and decision matrix
&lt;/h2&gt;

&lt;p&gt;The worst time to decide what constitutes a "failed migration" is at 2:00 AM when half the team wants to debug a broken service and the other half wants to abort. Human psychology works against you during a migration: because the team invested weeks of work into preparing the move, engineers naturally suffer from sunk-cost bias and will try to fix problems forward for too long.&lt;/p&gt;

&lt;p&gt;Protect your team by writing down hard, measurable rollback triggers before the migration starts. Designate one person as the Incident Commander who has the authority to call a rollback when a trigger is breached.&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Failure Scenario&lt;/th&gt;
&lt;th&gt;Metric / Trigger Threshold&lt;/th&gt;
&lt;th&gt;Action Required&lt;/th&gt;
&lt;th&gt;Phase&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Replication lag won't clear&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Lag remains &amp;gt; 5 seconds during write pause for more than 10 minutes&lt;/td&gt;
&lt;td&gt;Abort cutover; re-enable writes on old cloud&lt;/td&gt;
&lt;td&gt;Phase 1 (During cutover)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Smoke tests fail&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Critical path (login, checkout, API auth) fails for &amp;gt; 15 minutes after switch&lt;/td&gt;
&lt;td&gt;Abort cutover; switch DNS back to old cloud&lt;/td&gt;
&lt;td&gt;Phase 1 (During cutover)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Elevated HTTP 5xx errors&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;5xx error rate exceeds 2% for 15 consecutive minutes after fixing attempts&lt;/td&gt;
&lt;td&gt;Execute stateful rollback via reverse replication&lt;/td&gt;
&lt;td&gt;Phase 2 (Post-cutover)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Database saturation&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;CPU or disk I/O stays above 90% under normal load and scaling up fails&lt;/td&gt;
&lt;td&gt;Execute stateful rollback during next low-traffic window&lt;/td&gt;
&lt;td&gt;Phase 2 (Post-cutover)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Data corruption or loss&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Any verified instance of lost transactions or mismatched checksums&lt;/td&gt;
&lt;td&gt;Immediate emergency pause and rollback evaluation&lt;/td&gt;
&lt;td&gt;Phase 1 or Phase 2&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;h2&gt;
  
  
  Regulatory context: DORA and the EU Data Act
&lt;/h2&gt;

&lt;p&gt;For many European companies, having a tested rollback and exit plan is no longer just an engineering best practice—it is a legal requirement or a contractual obligation to enterprise customers.&lt;/p&gt;

&lt;p&gt;Since 17 January 2025, the &lt;strong&gt;Digital Operational Resilience Act (DORA, Regulation (EU) 2022/2554)&lt;/strong&gt; has applied to financial entities across the EU and their critical third-party ICT service providers. DORA requires covered organizations to maintain comprehensive ICT business continuity policies and documented exit strategies for critical functions. Crucially, regulators expect exit plans and migration procedures to be tested and to account for failed transitions without disrupting service delivery or losing data integrity.&lt;/p&gt;

&lt;p&gt;At the same time, the &lt;strong&gt;EU Data Act (Regulation (EU) 2023/2854)&lt;/strong&gt; makes keeping a rollback option financially easier. Historically, cloud providers charged high data egress fees whenever you replicated data out of their network. Under Article 29 of the EU Data Act, cloud switching charges—including data egress fees incurred for switching providers—are restricted to no more than the provider's direct cost until 12 January 2027, and are completely banned from 12 January 2027 onward. While you will still pay for the compute and storage resources while running two environments in parallel for a month, punitive network transfer fees for syncing data between providers are phasing out.&lt;/p&gt;

&lt;h2&gt;
  
  
  Pre-migration rollback checklist
&lt;/h2&gt;

&lt;p&gt;Before you approve your migration window, verify every item on this list:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;[ ] &lt;strong&gt;Old environment freeze:&lt;/strong&gt; Automatic teardown scripts, billing cleanup jobs, and lifecycle deletion rules on the source cloud are disabled for at least 30 days.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;DNS readiness:&lt;/strong&gt; TTLs on all production domains and subdomains have been lowered to 60–300 seconds at least 48 hours prior.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Access verification:&lt;/strong&gt; At least two engineers on the cutover call have working admin credentials and MFA access to both the old cloud console, the new cloud console, and the DNS registrar.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Write-pause mechanism:&lt;/strong&gt; A tested script or feature flag exists to block all user and background worker writes within 30 seconds.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Reverse replication tested:&lt;/strong&gt; Logical replication from the new database back to the old database has been tested in a staging rehearsal.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Object storage sync active:&lt;/strong&gt; A script or worker is ready to sync new S3/object storage files from the new cloud back to the old cloud.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Timebox agreed:&lt;/strong&gt; The team has agreed on a strict maximum debugging time (for example, 20 minutes during the cutover window) before triggering an automatic rollback.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Frequently asked questions
&lt;/h2&gt;

&lt;h3&gt;
  
  
  How long should we keep our old cloud environment running after migrating?
&lt;/h3&gt;

&lt;p&gt;For most production workloads, keeping the old environment running in a synced standby state for 14 to 30 days is the sweet spot. Two weeks covers weekly batch jobs and normal traffic cycles, while 30 days ensures you survive a full month-end billing or reporting cycle on the new infrastructure before deleting the old resources.&lt;/p&gt;

&lt;h3&gt;
  
  
  Does keeping a rollback environment double our cloud bill?
&lt;/h3&gt;

&lt;p&gt;It temporarily increases your infrastructure spend for the duration of the overlap, typically one month. However, because European cloud providers like Hetzner, UpCloud, Scaleway, and OVHcloud generally have lower baseline compute and bandwidth prices than US hyperscalers, running the new EU environment for a month alongside your old cloud usually pays for itself within a few months of shutting the old provider down. You can also scale down non-essential worker nodes in the old cloud while keeping the database warm.&lt;/p&gt;

&lt;h3&gt;
  
  
  Can we roll back if we change database engines during the migration?
&lt;/h3&gt;

&lt;p&gt;Rolling back across different database engines (for example, moving from AWS DynamoDB to PostgreSQL, or changing major database versions) is much harder because native logical replication requires compatible engines and schemas. Whenever possible, keep the exact same database engine and version during your cloud migration, and save database modernization projects for after you are stable on the new cloud.&lt;/p&gt;

&lt;h3&gt;
  
  
  What should we do with background queues during a cutover or rollback?
&lt;/h3&gt;

&lt;p&gt;Before pausing database writes, pause your queue consumers (such as Sidekiq, Celery, BullMQ, or RabbitMQ workers) and let in-flight jobs finish. Wait for the queues to drain completely before switching environments, or ensure unprocessed job payloads are replicated to the destination Redis or queue instance so jobs are not dropped or executed twice.&lt;/p&gt;

&lt;h2&gt;
  
  
  Sources
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;&lt;a href="https://eur-lex.europa.eu/eli/reg/2023/2854/oj" rel="noopener noreferrer"&gt;Regulation (EU) 2023/2854 (EU Data Act) on EUR-Lex&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://eur-lex.europa.eu/eli/reg/2022/2554/oj" rel="noopener noreferrer"&gt;Regulation (EU) 2022/2554 (Digital Operational Resilience Act - DORA) on EUR-Lex&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://www.postgresql.org/docs/current/logical-replication.html" rel="noopener noreferrer"&gt;PostgreSQL Official Documentation: Logical Replication&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://dev.mysql.com/doc/refman/8.0/en/replication-gtids.html" rel="noopener noreferrer"&gt;MySQL Official Documentation: Replication with Global Transaction Identifiers (GTIDs)&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;




&lt;p&gt;&lt;em&gt;Drafted with AI assistance and fact-checked by the MoveInfra team. The original, kept up to date, is at &lt;a href="https://moveinfra.eu/blog/rollback-plan-cloud-migration/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu&lt;/a&gt;. Free tools for pricing a move to an EU cloud: &lt;a href="https://moveinfra.eu/tools/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu/tools&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>cloud</category>
      <category>cloudmigration</category>
      <category>rollbackplan</category>
      <category>databasereplication</category>
    </item>
    <item>
      <title>What happens on switch day: cloud migration cutover explained</title>
      <dc:creator>Denys Zaruin</dc:creator>
      <pubDate>Wed, 04 Nov 2026 13:35:54 +0000</pubDate>
      <link>https://dev.to/moveinfra-eu/what-happens-on-switch-day-cloud-migration-cutover-explained-44fa</link>
      <guid>https://dev.to/moveinfra-eu/what-happens-on-switch-day-cloud-migration-cutover-explained-44fa</guid>
      <description>&lt;p&gt;&lt;em&gt;Originally published on &lt;a href="https://moveinfra.eu/blog/what-is-a-cloud-migration-cutover/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;A cloud migration cutover is the final phase of moving infrastructure from one cloud provider to another, where live user traffic and database writes are switched from the old environment to the new one. For most European small and mid-sized companies, a well-planned cutover takes between five and thirty minutes of planned maintenance. During this short window, your engineering team pauses new data writes on the old cloud, waits a few seconds for the new database to finish syncing, updates the Domain Name System (DNS) records to point to the new servers, and verifies that the application works properly before reopening traffic to users.&lt;/p&gt;

&lt;h2&gt;
  
  
  Key takeaways
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Cutover is the final step, not the whole migration:&lt;/strong&gt; Building servers, configuring networks, and copying bulk data happen weeks before switch day.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Brief write pauses beat complex active-active setups:&lt;/strong&gt; Accepting a few minutes of read-only mode or a maintenance page is much safer and cheaper than trying to write to two different clouds at the same time.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Preparation controls the clock:&lt;/strong&gt; Lowering DNS Time-to-Live (TTL) values days in advance and keeping databases in live sync reduces actual cutover time to minutes.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Always keep an exit path:&lt;/strong&gt; Reverse database replication back to the old cloud lets you undo the switch without losing new customer data if an unexpected bug appears days later.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  What is a cloud migration cutover?
&lt;/h2&gt;

&lt;p&gt;When companies talk about moving from a US hyperscaler like AWS or Google Cloud to a European cloud provider, they often picture a massive weekend project where files are copied over while the system is offline. In modern cloud operations, the bulk work happens quietly in the background weeks before anyone touches live traffic.&lt;/p&gt;

&lt;p&gt;You build the new virtual machines, Kubernetes clusters, load balancers, and firewalls using infrastructure-as-code tools like Terraform. You deploy your application code to the new servers through your regular deployment pipelines. You also set up a continuous data link so your new database and object storage buckets mirror everything happening in your old environment in real time.&lt;/p&gt;

&lt;p&gt;The &lt;strong&gt;cutover&lt;/strong&gt; is simply the moment you change which environment is in charge. Before the cutover, your old cloud is primary (handling live users) and your new cloud is a warm standby. After the cutover, your new cloud is primary, and your old cloud becomes the fallback.&lt;/p&gt;

&lt;h2&gt;
  
  
  Why a short write pause is the safest approach
&lt;/h2&gt;

&lt;p&gt;Vendors often market the idea of switching clouds with no interruption at all. While technically possible for some architectures, achieving uninterrupted writes across two different cloud providers requires an "active-active" database setup. This means both the old database (for example, on AWS RDS) and the new database (for example, on UpCloud or Scaleway) accept new customer data at the exact same moment.&lt;/p&gt;

&lt;p&gt;If two customers update the same record on different clouds within the same millisecond, you get a data conflict. Resolving these conflicts requires complex application changes, custom database routing, and expensive engineering time. For a company of 10 to 300 people, this complexity introduces far more risk than it removes.&lt;/p&gt;

&lt;p&gt;A &lt;strong&gt;planned write pause&lt;/strong&gt; is the practical alternative. You schedule the cutover during your lowest traffic window, such as Tuesday at 03:00 CET. You put your application into maintenance mode or read-only mode for a few minutes. Because no new data is entering the old database, the new database catches up completely within seconds. You then promote the new database to be the single source of truth and reopen the application on the new cloud. Your data remains completely consistent, and the engineering effort stays manageable.&lt;/p&gt;

&lt;h2&gt;
  
  
  What happens before switch day
&lt;/h2&gt;

&lt;p&gt;A calm cutover depends on three preparation steps completed at least 48 to 72 hours before the switch.&lt;/p&gt;

&lt;h3&gt;
  
  
  1. Lowering DNS TTLs
&lt;/h3&gt;

&lt;p&gt;The Domain Name System (DNS) is the phonebook of the internet, translating your domain (like &lt;code&gt;app.example.eu&lt;/code&gt;) into the IP address of your load balancer. Every DNS record has a Time-to-Live (TTL) setting, measured in seconds, which tells internet service providers how long to cache that IP address before checking for an update.&lt;/p&gt;

&lt;p&gt;Many domains have a default TTL of 3,600 seconds (one hour) or even 86,400 seconds (24 hours). If you change your DNS record while the TTL is set to 24 hours, some users will still be sent to your old cloud a full day later. Several days before the cutover, you lower the TTL on all relevant records to 60 or 300 seconds. That way, when you change the IP address on switch day, the rest of the internet notices the change within a few minutes.&lt;/p&gt;

&lt;h3&gt;
  
  
  2. Live data synchronization
&lt;/h3&gt;

&lt;p&gt;Instead of taking a static backup on cutover night and waiting hours for it to restore, teams use &lt;strong&gt;logical replication&lt;/strong&gt; for databases like PostgreSQL and MySQL. Logical replication copies the initial dataset and then streams every new insert, update, and delete from the old cloud to the new cloud as it happens. The time difference between the two databases is called &lt;strong&gt;replication lag&lt;/strong&gt;. Under normal conditions, this lag is less than a second.&lt;/p&gt;

&lt;p&gt;For object storage (like moving from Amazon S3 to an S3-compatible bucket on Hetzner or OVHcloud), you run an initial sync days in advance using tools that verify file integrity with cryptographic checksums. On cutover night, you only need to sync the small batch of files created in the final few minutes.&lt;/p&gt;

&lt;h3&gt;
  
  
  3. Dry runs and synthetic testing
&lt;/h3&gt;

&lt;p&gt;Before moving real users, your team tests the new environment privately. By overriding local DNS settings on their own computers (using the &lt;code&gt;/etc/hosts&lt;/code&gt; file) or using internal test domains, engineers can browse the application on the new cloud, verify SSL certificates, check firewall rules, and confirm that background worker queues and caches (such as Redis) respond as expected.&lt;/p&gt;

&lt;h2&gt;
  
  
  Step-by-step: the cutover timeline
&lt;/h2&gt;

&lt;p&gt;Below is what a typical 30-minute cutover window looks like for a standard web application using a relational database, object storage, and a load balancer.&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Time&lt;/th&gt;
&lt;th&gt;Phase&lt;/th&gt;
&lt;th&gt;Action&lt;/th&gt;
&lt;th&gt;System State&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;T-15 min&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Pre-flight check&lt;/td&gt;
&lt;td&gt;Verify new environment health, check SSL certificates, and confirm database replication lag is under 5 seconds.&lt;/td&gt;
&lt;td&gt;Live on old cloud&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;T-0 min&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Pause writes&lt;/td&gt;
&lt;td&gt;Enable maintenance page or read-only mode on the application. Stop background job processors (cron jobs, queue workers) on the old cloud.&lt;/td&gt;
&lt;td&gt;Maintenance mode&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;T+2 min&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Drain &amp;amp; sync&lt;/td&gt;
&lt;td&gt;Wait for in-flight database transactions to finish. Confirm replication lag reaches zero and run a final delta sync on object storage.&lt;/td&gt;
&lt;td&gt;Maintenance mode&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;T+5 min&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Promote target&lt;/td&gt;
&lt;td&gt;Stop replication. Promote the new EU cloud database from read-only replica to primary (read-write).&lt;/td&gt;
&lt;td&gt;Maintenance mode&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;T+7 min&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Start services&lt;/td&gt;
&lt;td&gt;Enable background job processors on the new cloud. Start reverse replication from the new database back to the old database (if using a warm rollback setup).&lt;/td&gt;
&lt;td&gt;Maintenance mode&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;T+10 min&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Switch traffic&lt;/td&gt;
&lt;td&gt;Update DNS records (A, AAAA, or CNAME) to point to the new cloud's load balancer IP. Disable the maintenance page on the new cloud.&lt;/td&gt;
&lt;td&gt;Live on new cloud&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;T+15 min&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Verify &amp;amp; monitor&lt;/td&gt;
&lt;td&gt;Run automated smoke tests (login, checkout, API writes). Monitor error logs, CPU usage, and database connections as DNS change propagates.&lt;/td&gt;
&lt;td&gt;Live on new cloud&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;h3&gt;
  
  
  A closer look at the critical steps
&lt;/h3&gt;

&lt;p&gt;&lt;strong&gt;Stopping background workers (T-0 min):&lt;/strong&gt; Web traffic is not the only thing that writes to your database. Scheduled cron jobs, billing scripts, and message queues also change data. If you block web users at the load balancer but forget to pause a background billing worker on the old cloud, that worker might create an invoice on the old database after you have already switched to the new one. Stopping all background processes is mandatory before promoting the new database.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Checking replication lag (T+2 min):&lt;/strong&gt; Once writes are stopped on the old cloud, engineers check the database metrics to verify that the new database has processed the final bytes of the replication stream. Because the databases were already in sync, this usually takes less than five seconds.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Handling stubborn DNS caches (T+10 min):&lt;/strong&gt; Even if you set your DNS TTL to 60 seconds, a small percentage of corporate networks and mobile carriers ignore TTL rules and cache records longer. Because you kept the old environment in maintenance mode (or closed its database ports), any user whose network sends them to the old IP address simply sees the maintenance page until their local DNS cache refreshes, rather than accidentally writing data to the old database. Alternatively, some teams configure the old load balancer to proxy traffic directly to the new load balancer during the transition period.&lt;/p&gt;

&lt;h2&gt;
  
  
  What can go wrong (and how to handle it)
&lt;/h2&gt;

&lt;p&gt;Even with thorough testing, production traffic behaves differently than test traffic. Here are the most common hiccups during a cutover and how operations teams resolve them:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Third-party IP allowlists:&lt;/strong&gt; If your application connects to external partners—such as payment gateways, bank APIs, or enterprise customer servers—those partners might block requests coming from your new cloud's outbound IP addresses. &lt;em&gt;Prevention:&lt;/em&gt; Audit outbound integrations beforehand and give partners your new static NAT gateway IPs a week early.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Email delivery blocks:&lt;/strong&gt; Sending transactional emails directly from new cloud servers often triggers spam filters because the new IP addresses have no sending reputation. If you are replacing a cloud-native email service like AWS SES, configure and verify domain records (SPF, DKIM, DMARC) with a dedicated email provider well before cutover night.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Hidden hardcoded configurations:&lt;/strong&gt; Occasionally, an environment variable or configuration file still points to an old internal IP address or proprietary cloud endpoint. Having your CI/CD pipeline ready during the cutover window allows your team to push a quick configuration fix in minutes.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Rollback: keeping your safety net
&lt;/h2&gt;

&lt;p&gt;For European companies—especially those subject to financial regulations like DORA (Regulation (EU) 2022/2554), which requires tested ICT exit strategies and safe transitions—having a reliable rollback plan is essential.&lt;/p&gt;

&lt;p&gt;Before you promote the new database (Step T+5 in the table above), rolling back is trivial: you simply turn off the maintenance page on the old cloud and cancel the window.&lt;/p&gt;

&lt;p&gt;Once you promote the new database and users start saving new data on the new cloud, rolling back requires more care. If you simply point DNS back to the old cloud three hours later, you would lose all orders, messages, or account changes made during those three hours.&lt;/p&gt;

&lt;p&gt;To prevent this, mature migration plans set up &lt;strong&gt;reverse replication&lt;/strong&gt; during the cutover. Once the new EU database becomes the primary, the old cloud database is reconfigured to act as a follower, receiving a live stream of all new changes happening on the EU cloud. If a severe issue appears two days after the cutover, you can repeat a mini-cutover in reverse—pausing writes, letting the old cloud catch up, and switching DNS back—without losing a single row of customer data.&lt;/p&gt;

&lt;p&gt;We recommend keeping the old environment running in this reverse-synced state for up to 30 days. Under the EU Data Act (Regulation (EU) 2023/2854), cloud providers may charge no more than their direct cost for switching-related data egress until 12 January 2027, after which switching egress fees are banned entirely across the EU. Once your team is completely confident in the new environment, you sever the replication link and delete the old cloud resources.&lt;/p&gt;

&lt;h2&gt;
  
  
  Frequently asked questions
&lt;/h2&gt;

&lt;h3&gt;
  
  
  How long does a cloud migration cutover take?
&lt;/h3&gt;

&lt;p&gt;For most small and mid-sized architectures where databases and storage are synced in advance, the cutover window takes between 5 and 30 minutes. The actual pause in database writes often lasts less than five minutes; the remaining time is spent running health checks and waiting for DNS changes to propagate.&lt;/p&gt;

&lt;h3&gt;
  
  
  Can we do a cutover during business hours?
&lt;/h3&gt;

&lt;p&gt;You can, especially if your system uses a load balancer proxy method to forward old traffic, or if your application can run in read-only mode without bothering users. However, most companies prefer to schedule the cutover during their quietest night or weekend hours so engineers can work without pressure and fewer users see the brief maintenance window.&lt;/p&gt;

&lt;h3&gt;
  
  
  What happens to active user sessions during the switch?
&lt;/h3&gt;

&lt;p&gt;If your application stores user login sessions inside a database or a Redis cache that is replicated to the new cloud, users will stay logged in after the switch. If sessions are stored only in the local memory of the old servers, or if you start with a fresh Redis cache on the new cloud, active users will need to log in again once traffic moves over.&lt;/p&gt;

&lt;h3&gt;
  
  
  When should we shut down the old cloud servers?
&lt;/h3&gt;

&lt;p&gt;Do not delete your old servers immediately after the cutover succeeds. Keep the old infrastructure running for at least one to four weeks while monitoring performance and monthly billing on the new provider. Once all weekly or monthly background jobs have run at least once on the new cloud without issues, you can safely decommission the old account.&lt;/p&gt;

&lt;h2&gt;
  
  
  Sources
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;&lt;a href="https://eur-lex.europa.eu/eli/reg/2023/2854/oj" rel="noopener noreferrer"&gt;Regulation (EU) 2023/2854 (EU Data Act) on EUR-Lex&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://eur-lex.europa.eu/eli/reg/2022/2554/oj" rel="noopener noreferrer"&gt;Regulation (EU) 2022/2554 (Digital Operational Resilience Act - DORA) on EUR-Lex&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://www.postgresql.org/docs/current/logical-replication.html" rel="noopener noreferrer"&gt;PostgreSQL Official Documentation: Logical Replication&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://datatracker.ietf.org/doc/html/rfc1035" rel="noopener noreferrer"&gt;RFC 1035: Domain Names - Implementation and Specification (IETF)&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;




&lt;p&gt;&lt;em&gt;Drafted with AI assistance and fact-checked by the MoveInfra team. The original, kept up to date, is at &lt;a href="https://moveinfra.eu/blog/what-is-a-cloud-migration-cutover/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu&lt;/a&gt;. Free tools for pricing a move to an EU cloud: &lt;a href="https://moveinfra.eu/tools/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu/tools&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>cloud</category>
      <category>cloudmigration</category>
      <category>cutover</category>
      <category>dns</category>
    </item>
    <item>
      <title>How to write a cloud exit plan (with a checklist)</title>
      <dc:creator>Denys Zaruin</dc:creator>
      <pubDate>Tue, 03 Nov 2026 13:25:07 +0000</pubDate>
      <link>https://dev.to/moveinfra-eu/how-to-write-a-cloud-exit-plan-with-a-checklist-359o</link>
      <guid>https://dev.to/moveinfra-eu/how-to-write-a-cloud-exit-plan-with-a-checklist-359o</guid>
      <description>&lt;p&gt;&lt;em&gt;Originally published on &lt;a href="https://moveinfra.eu/blog/cloud-exit-plan-template/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;To write a cloud exit plan, you need to document six things: what triggers an exit, a complete inventory of your cloud services and dependencies, a target provider mapping, a technical runbook for moving state and traffic, a timeline with named owners, and a way to test the plan at least once a year. A useful exit strategy is not a legal binder written to sit on a shelf. It is an operational runbook that explains how your company can move its infrastructure, data, and deployments from a current cloud provider (such as AWS, Google Cloud, or Azure) to an alternative provider within a known timeframe and cost budget, while keeping data intact and service interruptions predictable.&lt;/p&gt;

&lt;h2&gt;
  
  
  Key takeaways
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Regulators and enterprise customers increasingly expect written exit plans.&lt;/strong&gt; Under DORA (Regulation (EU) 2022/2554), financial entities and many of their technology suppliers must maintain tested exit strategies for critical ICT services.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;The EU Data Act changes the economics of switching.&lt;/strong&gt; Under Article 29 of Regulation (EU) 2023/2854, cloud switching charges—including data egress fees for switching—are capped at the provider's direct cost until 12 January 2027, and banned entirely from 12 January 2027.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Data gravity and proprietary services are the main bottlenecks.&lt;/strong&gt; Virtual machines and standard PostgreSQL databases move easily; proprietary NoSQL databases, identity rules, and undocumented network configurations take the longest to untangle.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;An untested plan is only a guess.&lt;/strong&gt; You do not need to migrate your entire production environment every year to validate your plan, but you do need to test restoring data and deploying your application to a secondary environment.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Why European companies need a real cloud exit plan
&lt;/h2&gt;

&lt;p&gt;For years, cloud exit plans were treated as theoretical exercises. Only heavily regulated banks wrote them, and engineers rarely read them. Today, three practical pressures make a real exit plan necessary for 10-to-300-person European companies.&lt;/p&gt;

&lt;p&gt;The first is regulatory compliance. The Digital Operational Resilience Act (DORA, Regulation (EU) 2022/2554) has applied since 17 January 2025. Under Article 28(8), financial entities must establish exit strategies for ICT services that support critical or important functions. Even if your company is not a licensed financial institution, selling software to European banks, insurers, or fintechs means their procurement teams will ask for your cloud exit documentation during vendor due diligence. Similarly, organizations subject to the NIS2 Directive (Directive (EU) 2022/2555) face stricter supply chain and business continuity expectations.&lt;/p&gt;

&lt;p&gt;The second is commercial leverage and cost control. Cloud pricing structures change, startup credit programs expire, and US-dollar billing introduces currency risk for Euro-denominated budgets. Having a documented, realistic path off your current hyperscaler turns contract renewals into genuine negotiations. Furthermore, the EU Data Act (Regulation (EU) 2023/2854), which became applicable in September 2025, mandates contractual and technical rules that make switching cloud providers easier. Specifically, Article 29 limits switching charges (such as data egress fees incurred when moving to another provider) to no more than the provider's actual cost until 12 January 2027, and bans those charges completely starting 12 January 2027.&lt;/p&gt;

&lt;p&gt;The third is sovereignty and jurisdictional risk. Many European boards and customers want to reduce reliance on non-EU cloud providers subject to US extraterritorial laws. Note that while offerings like the AWS European Sovereign Cloud keep data and operations inside the European Union, the provider remains Amazon, a US company. If your board or a major public-sector customer requires moving to an EU-owned cloud provider (such as UpCloud, Scaleway, Hetzner, OVHcloud, STACKIT, IONOS, or Exoscale), an existing exit plan saves weeks of initial discovery.&lt;/p&gt;

&lt;h2&gt;
  
  
  The six components of a practical cloud exit plan
&lt;/h2&gt;

&lt;p&gt;A readable cloud exit plan should be five to ten pages long, plus links to living technical inventories (like your Infrastructure as Code repository). Structure your document around the following six sections.&lt;/p&gt;

&lt;h3&gt;
  
  
  1. Scope, triggers, and governance
&lt;/h3&gt;

&lt;p&gt;Start by defining what the plan covers and when you would actually use it. Do not write a vague statement about "business continuity." List specific, measurable trigger events. Common triggers include:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Commercial triggers:&lt;/strong&gt; A price increase above a certain percentage, or the expiration of a committed-use discount where renewal terms are unacceptable.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Regulatory or customer triggers:&lt;/strong&gt; A mandate from a regulator or a key customer requiring infrastructure to be hosted on an EU-owned cloud.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Operational triggers:&lt;/strong&gt; Repeated multi-hour outages in your primary region, or termination of a specific cloud service your product relies on.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Legal triggers:&lt;/strong&gt; Changes in data transfer laws, sanctions, or insolvency of a provider.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Assign two clear roles: an &lt;strong&gt;Executive Owner&lt;/strong&gt; (usually the CEO, CFO, or COO) who has the authority to approve the budget and trigger the exit, and a &lt;strong&gt;Technical Lead&lt;/strong&gt; (CTO, VP of Engineering, or Lead DevOps Engineer) responsible for keeping the technical inventory accurate and running the migration.&lt;/p&gt;

&lt;h3&gt;
  
  
  2. Asset and dependency inventory
&lt;/h3&gt;

&lt;p&gt;You cannot move what you have not mapped. Most teams know where their main application servers and primary database live, but production cutovers fail because of forgotten background workers, hardcoded IP addresses, or managed security policies.&lt;/p&gt;

&lt;p&gt;Create a complete inventory of your current cloud footprint. If you use Terraform, OpenTofu, or Pulumi, your state files are a strong starting point, though you should still check the cloud billing console for resources created manually in the UI ("ClickOps"). Group every resource into one of three portability tiers:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;Standard and portable:&lt;/strong&gt; Linux virtual machines, Docker containers, standard Kubernetes workloads, managed PostgreSQL or MySQL databases, Redis-compatible caches, and S3-compatible object storage.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Moderate coupling (requires refactoring or configuration translation):&lt;/strong&gt; Cloud-specific load balancer annotations, managed message queues (like AWS SQS or Google Cloud Pub/Sub), serverless functions (AWS Lambda), and managed DNS (Route 53).&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;High lock-in (requires architectural redesign):&lt;/strong&gt; Proprietary NoSQL databases (DynamoDB, Firestore, Cosmos DB), proprietary data warehouses (BigQuery, Redshift, Synapse), and deep integrations with cloud-native identity systems (IAM roles baked into application logic).&lt;/li&gt;
&lt;/ol&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Component Category&lt;/th&gt;
&lt;th&gt;Typical Hyperscaler Service&lt;/th&gt;
&lt;th&gt;EU Cloud Equivalent (e.g., UpCloud, Scaleway, OVHcloud)&lt;/th&gt;
&lt;th&gt;Exit Effort&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Compute &amp;amp; Containers&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;EC2, EKS, GKE, AKS&lt;/td&gt;
&lt;td&gt;Managed Kubernetes or standard Linux VMs&lt;/td&gt;
&lt;td&gt;Low (if containerized)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Relational Databases&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;RDS, Aurora, Cloud SQL&lt;/td&gt;
&lt;td&gt;Managed PostgreSQL / MySQL (or self-hosted on VMs)&lt;/td&gt;
&lt;td&gt;Low to Medium (depends on data size)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Caching&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;ElastiCache, Memorystore&lt;/td&gt;
&lt;td&gt;Managed Redis / Valkey&lt;/td&gt;
&lt;td&gt;Low&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Object Storage&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Amazon S3, Cloud Storage&lt;/td&gt;
&lt;td&gt;S3-compatible Object Storage&lt;/td&gt;
&lt;td&gt;Low (requires checksum verification)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Queues &amp;amp; Serverless&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;SQS, Lambda, Cloud Functions&lt;/td&gt;
&lt;td&gt;RabbitMQ / Kafka on VMs, or provider serverless offerings&lt;/td&gt;
&lt;td&gt;Medium (requires code/config changes)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Proprietary Data&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;DynamoDB, BigQuery, Firestore&lt;/td&gt;
&lt;td&gt;Open-source alternatives (ClickHouse, MongoDB, Postgres)&lt;/td&gt;
&lt;td&gt;High (requires schema &amp;amp; app rewrite)&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;h3&gt;
  
  
  3. Target provider selection and mapping
&lt;/h3&gt;

&lt;p&gt;An exit plan must name a specific alternative provider, or at least a primary and secondary candidate. Writing "we will move to another cloud" hides technical incompatibilities.&lt;/p&gt;

&lt;p&gt;When choosing an EU-owned cloud target, match the provider's product catalog to your team's operational skills:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;If your team relies on managed databases, managed Kubernetes, and S3-compatible object storage out of the box, providers like &lt;strong&gt;UpCloud&lt;/strong&gt;, &lt;strong&gt;Scaleway&lt;/strong&gt;, &lt;strong&gt;OVHcloud&lt;/strong&gt;, &lt;strong&gt;STACKIT&lt;/strong&gt;, &lt;strong&gt;IONOS&lt;/strong&gt;, and &lt;strong&gt;Exoscale&lt;/strong&gt; offer managed equivalents for core primitives.&lt;/li&gt;
&lt;li&gt;If your team prefers raw price-performance and is comfortable managing software on bare metal or standard virtual machines, &lt;strong&gt;Hetzner&lt;/strong&gt; is a popular choice. However, your exit plan must note that Hetzner does not offer managed databases or managed Kubernetes, meaning your team (or your migration tooling) will need to configure and maintain PostgreSQL, MySQL, and Kubernetes on virtual servers directly.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Document the exact instance sizes, storage classes, and network topologies you will use on the target provider, along with an estimated monthly run-rate cost based on public list prices.&lt;/p&gt;

&lt;h3&gt;
  
  
  4. Data migration and synchronization strategy
&lt;/h3&gt;

&lt;p&gt;Code is stateless and easy to redeploy; data has gravity. Your exit plan must explain how you will move gigabytes or terabytes of state without corrupting records or taking your business offline for days.&lt;/p&gt;

&lt;p&gt;Break your data strategy down by storage type:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Relational databases (PostgreSQL, MySQL):&lt;/strong&gt; Taking a static &lt;code&gt;pg_dump&lt;/code&gt; or snapshot and restoring it on the new cloud works only if your database is small and your business can afford several hours of downtime. For most production systems, specify &lt;strong&gt;logical replication&lt;/strong&gt;. You set up the target database on the new cloud, copy the initial snapshot, and stream ongoing writes in real time until the new database is less than a few seconds behind the source.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Object storage (S3, GCS):&lt;/strong&gt; Specify a tool (such as &lt;code&gt;rclone&lt;/code&gt;) that copies buckets using standard S3 APIs. Because object buckets often hold millions of user files or invoices, your plan must require checksum verification on all transferred files and continuous delta syncs leading up to the cutover date.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Encryption keys and secrets:&lt;/strong&gt; Document how TLS certificates, database credentials, and encryption keys stored in services like AWS KMS or Secrets Manager will be exported and injected into the target environment.&lt;/li&gt;
&lt;/ul&gt;

&lt;h3&gt;
  
  
  5. Cutover, rollback, and cleanup runbook
&lt;/h3&gt;

&lt;p&gt;The cutover section is the step-by-step sequence your engineering team will follow on migration day. Keep it chronological and explicit:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;Pre-cutover preparation (24–48 hours prior):&lt;/strong&gt; Lower DNS Time-To-Live (TTL) values to 60 or 300 seconds so DNS changes propagate quickly across the internet. Verify that database logical replication lag is consistently near zero and object storage delta syncs are complete.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Deployment pipeline alignment:&lt;/strong&gt; Update your CI/CD pipelines (GitHub Actions, GitLab CI, etc.) so your team can deploy the exact same application build to both the old and new environments.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;The switch window:&lt;/strong&gt; Put the application into maintenance mode or pause database writes on the old cloud. Wait until replication lag reaches zero (typically under 5 seconds). Run automated health and data integrity checks on the target database.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;DNS cutover:&lt;/strong&gt; Point your DNS records (or external load balancer) to the new EU cloud environment and enable writes.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Rollback window (Day 1 to Day 30):&lt;/strong&gt; Never delete the old cloud environment immediately after switching DNS. A robust exit plan keeps the old environment alive and sets up &lt;strong&gt;reverse replication&lt;/strong&gt;—streaming new writes from the new EU cloud back to the old cloud for 14 to 30 days. If an unexpected network or performance issue appears in the new environment, you can switch DNS back without losing data written after the migration.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Decommissioning:&lt;/strong&gt; Once the rollback period passes without critical issues, stop reverse replication, take a final archival backup, delete data on the old cloud according to GDPR and contractual rules, and close the account.&lt;/li&gt;
&lt;/ol&gt;

&lt;h3&gt;
  
  
  6. Testing, maintenance, and cost estimation
&lt;/h3&gt;

&lt;p&gt;Under DORA and standard ISO 27001 business continuity rules, an exit plan must be reviewed and tested regularly. Include a schedule in your document:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Every 6 months:&lt;/strong&gt; Update the resource inventory, verify that new cloud services introduced by product teams are logged in the portability table, and check target cloud pricing.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Every 12 months:&lt;/strong&gt; Run a tabletop exercise with engineering and ops leads, or perform a technical dry run where you spin up the target infrastructure from your Terraform code, restore a database backup, and deploy the staging application.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Finally, include a realistic budget estimate for executing the exit. Break costs into three buckets:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;Parallel run costs:&lt;/strong&gt; Paying for both the old cloud and the new target cloud simultaneously during the testing, replication, and 30-day rollback window (typically 1 to 2 months of duplicate infrastructure spend).&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Switching and egress fees:&lt;/strong&gt; Note that under EU Data Act Article 29, providers may only charge at-cost fees for data egress related to switching until 12 January 2027, and €0 after that date. However, you should check your current provider's specific waiver request process, as hyperscalers often require you to open a support ticket before migrating to qualify for reduced or waived egress fees.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Engineering time or external tooling:&lt;/strong&gt; Estimate the internal staff hours required to rewrite cloud-specific configurations, or the fixed cost of using an automated migration service.&lt;/li&gt;
&lt;/ol&gt;

&lt;h2&gt;
  
  
  Cloud exit plan checklist
&lt;/h2&gt;

&lt;p&gt;Use this checklist to audit your draft plan before submitting it to your board, auditor, or enterprise customer.&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;[ ] &lt;strong&gt;Triggers defined:&lt;/strong&gt; Specific commercial, legal, or operational events that start the exit process are listed.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Owners assigned:&lt;/strong&gt; Both an Executive Owner (budget authority) and Technical Lead (execution authority) are named.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Contract terms reviewed:&lt;/strong&gt; Notice periods, minimum spend commitments, and EU Data Act switching fee request procedures are documented.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Full inventory attached:&lt;/strong&gt; All compute, database, storage, network, and third-party SaaS integrations are listed.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Lock-in risks flagged:&lt;/strong&gt; Proprietary services (e.g., DynamoDB, SQS, Lambda, IAM roles) are isolated with a clear refactoring or replacement plan.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Target cloud mapped:&lt;/strong&gt; A specific alternative provider is chosen, with matching instance sizes, regions, and estimated monthly costs.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Infrastructure as Code ready:&lt;/strong&gt; Network topologies, firewalls, and compute resources can be provisioned reproducibly without manual console clicks.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Data sync method tested:&lt;/strong&gt; Live database replication (e.g., logical replication) and object storage cloning with checksums are documented.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;CI/CD pipeline documented:&lt;/strong&gt; Steps required to point automated code deployments to the new environment are written down.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Rollback mechanism included:&lt;/strong&gt; The plan keeps the old environment active with reverse data replication for a defined safety period (e.g., 30 days) before deletion.&lt;/li&gt;
&lt;li&gt;[ ] &lt;strong&gt;Review cadence set:&lt;/strong&gt; A recurring calendar reminder is set to update the inventory and test the plan annually.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Frequently asked questions
&lt;/h2&gt;

&lt;h3&gt;
  
  
  Does DORA require a cloud exit plan for SaaS companies that are not banks?
&lt;/h3&gt;

&lt;p&gt;DORA applies directly to regulated financial entities in the EU, but those entities are legally required to manage third-party ICT risk across their supply chain. If you provide critical software or infrastructure to banks, insurers, or investment firms, their compliance teams will usually require you to demonstrate a documented business continuity and cloud exit strategy as part of your contract.&lt;/p&gt;

&lt;h3&gt;
  
  
  Can we still be charged egress fees when leaving AWS, Google Cloud, or Azure?
&lt;/h3&gt;

&lt;p&gt;Under Article 29 of the EU Data Act (Regulation (EU) 2023/2854), cloud providers cannot charge switching fees—including data egress fees for moving to another provider—that exceed their direct technical cost of facilitating the switch. Starting 12 January 2027, all switching and egress charges for changing providers are completely banned in the EU. Keep in mind that you usually must notify your current provider in advance and follow their documented switching procedure rather than simply downloading data without notice.&lt;/p&gt;

&lt;h3&gt;
  
  
  How long should a realistic cloud exit take for a 50-person company?
&lt;/h3&gt;

&lt;p&gt;If your infrastructure relies on standard workloads—such as Linux VMs, Kubernetes, PostgreSQL/MySQL, and S3-compatible object storage—executing a well-prepared exit plan typically takes two to four weeks of setup and synchronization, followed by a short planned cutover window and a 30-day rollback period. If your application is tightly coupled to proprietary services like DynamoDB or hundreds of serverless functions, refactoring the application code can add two to six months to the timeline.&lt;/p&gt;

&lt;h3&gt;
  
  
  Should we build a multi-cloud active-active architecture to make exiting easier?
&lt;/h3&gt;

&lt;p&gt;For most 10-to-300-person companies, running an active-active architecture across two cloud providers simultaneously is too expensive and operationally complex. It doubles your networking and monitoring overhead and introduces cross-cloud latency. A much more practical approach is &lt;strong&gt;portable single-cloud&lt;/strong&gt;: run on one provider at a time, use standard open-source databases and containers, keep your infrastructure in Terraform, and maintain a tested exit plan so you can move when necessary.&lt;/p&gt;

&lt;h2&gt;
  
  
  Sources
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;&lt;a href="https://eur-lex.europa.eu/eli/reg/2022/2554/oj" rel="noopener noreferrer"&gt;Regulation (EU) 2022/2554 (Digital Operational Resilience Act / DORA) on EUR-Lex&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://eur-lex.europa.eu/eli/reg/2023/2854/oj" rel="noopener noreferrer"&gt;Regulation (EU) 2023/2854 (EU Data Act) on EUR-Lex&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://eur-lex.europa.eu/eli/dir/2022/2555/oj" rel="noopener noreferrer"&gt;Directive (EU) 2022/2555 (NIS2 Directive) on EUR-Lex&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://www.enisa.europa.eu/publications/cloud-security-guide-for-smes" rel="noopener noreferrer"&gt;ENISA Cloud Security Guide for SMEs&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;




&lt;p&gt;&lt;em&gt;Drafted with AI assistance and fact-checked by the MoveInfra team. The original, kept up to date, is at &lt;a href="https://moveinfra.eu/blog/cloud-exit-plan-template/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu&lt;/a&gt;. Free tools for pricing a move to an EU cloud: &lt;a href="https://moveinfra.eu/tools/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu/tools&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>cloud</category>
      <category>cloudexitplan</category>
      <category>dora</category>
      <category>eudataact</category>
    </item>
    <item>
      <title>DigitalOcean to UpCloud, Scaleway or Hetzner: a simple move</title>
      <dc:creator>Denys Zaruin</dc:creator>
      <pubDate>Tue, 03 Nov 2026 08:24:12 +0000</pubDate>
      <link>https://dev.to/moveinfra-eu/digitalocean-to-upcloud-scaleway-or-hetzner-a-simple-move-c6j</link>
      <guid>https://dev.to/moveinfra-eu/digitalocean-to-upcloud-scaleway-or-hetzner-a-simple-move-c6j</guid>
      <description>&lt;p&gt;&lt;em&gt;Originally published on &lt;a href="https://moveinfra.eu/blog/digitalocean-to-eu-cloud/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;Moving from DigitalOcean to a European cloud provider like UpCloud, Scaleway, or Hetzner is one of the most straightforward infrastructure projects a small engineering team can undertake. Unlike the big US hyperscalers, DigitalOcean is built around standard, open-source primitives: Linux virtual machines, standard PostgreSQL and MySQL databases, S3-compatible object storage, and upstream Kubernetes. Because your applications are rarely tied to proprietary APIs, you do not need to rewrite your software to leave. You can replicate your setup on an EU-owned provider, synchronize your data over the network, and switch your traffic over a single planned maintenance window.&lt;/p&gt;

&lt;h2&gt;
  
  
  Key takeaways
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Low technical lock-in:&lt;/strong&gt; DigitalOcean Droplets, Managed Databases, and Spaces map directly to standard services on European clouds without code changes.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Three distinct EU targets:&lt;/strong&gt; UpCloud offers a very similar developer experience with high-performance storage; Scaleway provides the broadest ecosystem of managed services and serverless tools; Hetzner delivers the lowest compute prices if you are willing to manage databases and Kubernetes yourself.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;App Platform requires a choice:&lt;/strong&gt; If you use DigitalOcean App Platform, you will either move to Scaleway Serverless Containers or transition to a standard container setup using Docker or Kubernetes.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Egress costs are manageable:&lt;/strong&gt; Most DigitalOcean plans include a generous outbound data transfer pool, and under the EU Data Act, switching egress fees are capped at provider cost until January 12, 2027, when they are banned entirely.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Why leaving DigitalOcean is simpler than leaving a hyperscaler
&lt;/h2&gt;

&lt;p&gt;When companies migrate away from AWS, Google Cloud, or Azure, they spend most of their time untangling proprietary services. An application built on AWS often relies on IAM roles for internal authentication, DynamoDB for data storage, SQS for queuing, and Lambda for compute. Moving that application requires rewriting code and redesigning how services talk to each other.&lt;/p&gt;

&lt;p&gt;DigitalOcean was designed differently. It became popular with startups and small teams specifically because it avoided that complexity. When you deploy on DigitalOcean, you are usually using standard Linux virtual machines (Droplets), standard open-source databases (PostgreSQL, MySQL, Redis/Valkey), an S3-compatible object store (Spaces), and standard Kubernetes (DOKS).&lt;/p&gt;

&lt;p&gt;For a European company looking to bring its infrastructure under EU jurisdiction, this architecture is an advantage. Your application code speaks standard protocols: TCP/IP, SQL, Redis protocol, and the S3 HTTP API. European cloud providers speak these exact same protocols. &lt;/p&gt;

&lt;p&gt;Instead of an application refactoring project, a move from DigitalOcean to Europe is a pure infrastructure migration. You build an equivalent environment on the new provider, copy your data, update your deployment pipelines, and switch your DNS records.&lt;/p&gt;

&lt;h2&gt;
  
  
  Choosing your destination: UpCloud, Scaleway, or Hetzner
&lt;/h2&gt;

&lt;p&gt;European cloud infrastructure has matured significantly. For teams leaving DigitalOcean, three providers stand out depending on your budget, your reliance on managed services, and how much operations work your team wants to do.&lt;/p&gt;

&lt;h3&gt;
  
  
  UpCloud: the closest match for DigitalOcean teams
&lt;/h3&gt;

&lt;p&gt;Headquartered in Finland, UpCloud is built for small and mid-sized engineering teams that want high reliability, predictable Euro billing, and a clean control panel without dozens of obscure services.&lt;/p&gt;

&lt;p&gt;If your DigitalOcean setup consists of Droplets, Managed PostgreSQL or MySQL, Managed Redis/Valkey, Load Balancers, Managed Kubernetes, and Spaces, UpCloud has a direct one-to-one equivalent for every component. UpCloud is also well-known for its in-house MaxIOPS block storage, which offers noticeably faster disk read and write speeds than standard shared VPS storage. &lt;/p&gt;

&lt;p&gt;For pricing, UpCloud offers entry-level Starter plans starting at approximately €3 per month for 1 vCPU and 1 GB of RAM as of 2026, alongside Cloud Native and High Performance tiers (check UpCloud's pricing page for current specifications and regional availability). UpCloud also includes zero-cost data egress under a fair transfer policy across its plans.&lt;/p&gt;

&lt;h3&gt;
  
  
  Scaleway: the broadest managed ecosystem
&lt;/h3&gt;

&lt;p&gt;Based in France and part of the Iliad Group, Scaleway is the closest European equivalent to a full-featured cloud platform. Like UpCloud, it covers all the core DigitalOcean primitives: Virtual Instances, Managed Databases for PostgreSQL and MySQL, Managed Redis, Kubernetes Kapsule, Load Balancers, and S3-compatible Object Storage.&lt;/p&gt;

&lt;p&gt;Scaleway stands out if your team uses higher-level platform services. If you rely on DigitalOcean Functions or DigitalOcean App Platform, Scaleway offers Serverless Functions, Serverless Containers, and Serverless Jobs. It also provides managed messaging and queuing systems, transactional email (TEM), and an integrated observability platform (Cockpit). If your team wants to minimize server maintenance and rely on managed platform tools, Scaleway is the strongest fit.&lt;/p&gt;

&lt;h3&gt;
  
  
  Hetzner: maximum compute value for hands-on teams
&lt;/h3&gt;

&lt;p&gt;Based in Germany with data centers in Germany and Finland, Hetzner is famous across Europe for its price-to-performance ratio. If your primary goal is reducing your monthly infrastructure bill while keeping data in the EU, Hetzner is hard to beat.&lt;/p&gt;

&lt;p&gt;However, there is a clear trade-off: Hetzner focuses strictly on core infrastructure. It provides Cloud Servers (VMs), Block Storage Volumes, Private Networks, Firewalls, Load Balancers, and S3-compatible Object Storage. &lt;strong&gt;Hetzner does not offer managed databases, managed Redis, or managed Kubernetes.&lt;/strong&gt;&lt;/p&gt;

&lt;p&gt;If you move from DigitalOcean to Hetzner, any managed PostgreSQL clusters or DOKS (Kubernetes) clusters you currently run will need to be set up and maintained on Linux virtual machines. For teams with experienced DevOps or platform engineers, running PostgreSQL with automated backups or deploying a lightweight Kubernetes distribution (like k3s) on Hetzner VMs is routine. For teams without dedicated operations staff, the money saved on compute can easily be lost in engineering time spent managing database failovers and cluster upgrades.&lt;/p&gt;

&lt;h2&gt;
  
  
  Service-by-service mapping
&lt;/h2&gt;

&lt;p&gt;The table below shows how DigitalOcean products map to UpCloud, Scaleway, and Hetzner.&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;DigitalOcean Service&lt;/th&gt;
&lt;th&gt;UpCloud Equivalent&lt;/th&gt;
&lt;th&gt;Scaleway Equivalent&lt;/th&gt;
&lt;th&gt;Hetzner Equivalent&lt;/th&gt;
&lt;th&gt;Migration Complexity&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;
&lt;strong&gt;Droplets&lt;/strong&gt; (Shared &amp;amp; Dedicated)&lt;/td&gt;
&lt;td&gt;Cloud Servers (Starter, Cloud Native, High Performance)&lt;/td&gt;
&lt;td&gt;Virtual Instances (Development, General Purpose)&lt;/td&gt;
&lt;td&gt;Cloud Servers (Shared &amp;amp; Dedicated vCPU)&lt;/td&gt;
&lt;td&gt;
&lt;strong&gt;Low.&lt;/strong&gt; Standard Linux VMs. Rebuild via Terraform or Ansible.&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Managed PostgreSQL / MySQL&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Managed Databases (PostgreSQL, MySQL)&lt;/td&gt;
&lt;td&gt;Managed Databases for PostgreSQL and MySQL&lt;/td&gt;
&lt;td&gt;
&lt;em&gt;No managed service.&lt;/em&gt; Run on Cloud Servers (VMs).&lt;/td&gt;
&lt;td&gt;
&lt;strong&gt;Low&lt;/strong&gt; on UpCloud/Scaleway via logical replication. &lt;strong&gt;Medium&lt;/strong&gt; on Hetzner (requires self-hosting setup).&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Managed Caching (Redis / Valkey)&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Managed Valkey&lt;/td&gt;
&lt;td&gt;Managed Inference / Redis clusters&lt;/td&gt;
&lt;td&gt;
&lt;em&gt;No managed service.&lt;/em&gt; Run on Cloud Servers.&lt;/td&gt;
&lt;td&gt;
&lt;strong&gt;Low.&lt;/strong&gt; Often migrated by warming a new cache or syncing via replicaof.&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Spaces Object Storage&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;UpCloud Object Storage&lt;/td&gt;
&lt;td&gt;Scaleway Object Storage&lt;/td&gt;
&lt;td&gt;Hetzner Object Storage&lt;/td&gt;
&lt;td&gt;
&lt;strong&gt;Low.&lt;/strong&gt; All use the S3 API. Sync files with &lt;code&gt;rclone&lt;/code&gt; and update endpoint URLs.&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Kubernetes (DOKS)&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Managed Kubernetes (UKS)&lt;/td&gt;
&lt;td&gt;Kubernetes Kapsule / Kosmos&lt;/td&gt;
&lt;td&gt;
&lt;em&gt;No managed service.&lt;/em&gt; Run k3s/Talos on VMs.&lt;/td&gt;
&lt;td&gt;
&lt;strong&gt;Low&lt;/strong&gt; on UpCloud/Scaleway. Redeploy Helm charts and manifests. &lt;strong&gt;Medium&lt;/strong&gt; on Hetzner.&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Load Balancers&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;Managed LoadBalancer&lt;/td&gt;
&lt;td&gt;Load Balancers&lt;/td&gt;
&lt;td&gt;Cloud Load Balancers&lt;/td&gt;
&lt;td&gt;
&lt;strong&gt;Low.&lt;/strong&gt; Recreate listeners, health checks, and TLS certificates.&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;VPC &amp;amp; Cloud Firewalls&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;SDN Private Networks &amp;amp; Firewall&lt;/td&gt;
&lt;td&gt;VPC, Private Networks &amp;amp; Security Groups&lt;/td&gt;
&lt;td&gt;Networks &amp;amp; Cloud Firewalls&lt;/td&gt;
&lt;td&gt;
&lt;strong&gt;Low.&lt;/strong&gt; Translate IP ranges and port rules in Terraform.&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;Volumes Block Storage&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;MaxIOPS / Standard Block Storage&lt;/td&gt;
&lt;td&gt;Block Storage (5K / 15K IOPS)&lt;/td&gt;
&lt;td&gt;Cloud Volumes&lt;/td&gt;
&lt;td&gt;
&lt;strong&gt;Low.&lt;/strong&gt; Attach new volumes and sync filesystem data with &lt;code&gt;rsync&lt;/code&gt;.&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;&lt;strong&gt;App Platform / Functions&lt;/strong&gt;&lt;/td&gt;
&lt;td&gt;
&lt;em&gt;No direct PaaS.&lt;/em&gt; Run on Kubernetes or Docker VMs.&lt;/td&gt;
&lt;td&gt;Serverless Containers &amp;amp; Serverless Functions&lt;/td&gt;
&lt;td&gt;
&lt;em&gt;No direct PaaS.&lt;/em&gt; Run Coolify, Kamal, or Docker on VMs.&lt;/td&gt;
&lt;td&gt;
&lt;strong&gt;Medium.&lt;/strong&gt; Requires packaging apps as standard containers if not already containerized.&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;h2&gt;
  
  
  What about pricing and switching costs?
&lt;/h2&gt;

&lt;p&gt;Small teams often choose DigitalOcean for its predictable monthly pricing, where shared CPU Droplets start at approximately $4 to $7 per month and Spaces starts at a $5 per month base subscription as of 2026 (always check the provider's official pricing page for current rates).&lt;/p&gt;

&lt;p&gt;When moving to an EU cloud, you gain two financial benefits:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;Billing in Euros:&lt;/strong&gt; For European companies, paying in USD introduces foreign exchange fees and monthly accounting variance. UpCloud, Scaleway, and Hetzner all bill natively in Euros.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Lower or comparable resource costs:&lt;/strong&gt; European providers are generally priced very competitively against US counterparts. Hetzner in particular often reduces raw compute and storage bills substantially, while UpCloud and Scaleway offer comparable or better performance per Euro spent, especially when factoring in network transfer.&lt;/li&gt;
&lt;/ol&gt;

&lt;h3&gt;
  
  
  Egress fees and the EU Data Act
&lt;/h3&gt;

&lt;p&gt;During a migration, you have to copy your databases and object storage over the internet to the new cloud provider. On some clouds, this outbound data transfer (egress) causes a painful final bill.&lt;/p&gt;

&lt;p&gt;On DigitalOcean, this is rarely an issue for small and mid-sized teams. Droplets and Spaces plans include generous monthly outbound transfer allowances pooled across your account. Unless you are moving many terabytes of media files in a single month, your migration traffic will likely fit inside your existing included bandwidth pool.&lt;/p&gt;

&lt;p&gt;Furthermore, European law protects companies switching cloud providers. Under Article 29 of the EU Data Act (Regulation (EU) 2023/2854), cloud providers may not charge switching fees—including data egress charges incurred to move to another provider—above their actual direct cost, and from January 12, 2027, all switching charges and switching egress fees are banned entirely across the EU.&lt;/p&gt;

&lt;h2&gt;
  
  
  How to execute the migration: a 6-step checklist
&lt;/h2&gt;

&lt;p&gt;If you decide to move your DigitalOcean environment to UpCloud, Scaleway, or Hetzner, you do not need a six-month roadmap. Most small teams can complete the preparation in a few days and execute the switch during a quiet evening window.&lt;/p&gt;

&lt;h3&gt;
  
  
  1. Inventory your DigitalOcean account
&lt;/h3&gt;

&lt;ul&gt;
&lt;li&gt;List all active Droplets, their CPU/RAM sizes, and any local data stored on their disks (outside of databases).&lt;/li&gt;
&lt;li&gt;Check your Managed Databases: note the exact engine version (for example, PostgreSQL 15 or MySQL 8), database size, and extensions used.&lt;/li&gt;
&lt;li&gt;Check Spaces buckets: note total storage size, public vs. private access permissions (ACLs), and whether you use the built-in Spaces CDN.&lt;/li&gt;
&lt;li&gt;Document all Cloud Firewall rules, VPC subnet ranges, and Load Balancer health check settings.&lt;/li&gt;
&lt;li&gt;Lower the TTL (Time-To-Live) on your DNS records to 60 or 300 seconds a few days before the cutover.&lt;/li&gt;
&lt;/ul&gt;

&lt;h3&gt;
  
  
  2. Provision the EU target environment as code
&lt;/h3&gt;

&lt;ul&gt;
&lt;li&gt;Even if you created your DigitalOcean resources by clicking around in the web console, use the migration as an opportunity to write Terraform (or OpenTofu) configurations for your new EU cloud. UpCloud, Scaleway, and Hetzner all maintain official, well-documented Terraform providers.&lt;/li&gt;
&lt;li&gt;Create the private network, firewall rules, virtual machines, load balancers, and database clusters in your chosen EU region (such as Frankfurt, Amsterdam, Paris, Warsaw, or Helsinki).&lt;/li&gt;
&lt;li&gt;Do not copy old virtual machine disk snapshots directly unless necessary. Instead, provision fresh Linux VMs on the EU cloud and deploy your application code to them.&lt;/li&gt;
&lt;/ul&gt;

&lt;h3&gt;
  
  
  3. Synchronize your object storage
&lt;/h3&gt;

&lt;ul&gt;
&lt;li&gt;Create the corresponding buckets on UpCloud, Scaleway, or Hetzner Object Storage.&lt;/li&gt;
&lt;li&gt;Use a tool like &lt;code&gt;rclone&lt;/code&gt; configured with both your DigitalOcean Spaces keys and your new EU S3 keys.&lt;/li&gt;
&lt;li&gt;Run an initial background sync (&lt;code&gt;rclone sync&lt;/code&gt; with checksum verification). Because this runs while your live application is still on DigitalOcean, it can take several hours or days without affecting users.&lt;/li&gt;
&lt;li&gt;Just before cutover, you will run a final delta sync to catch any newly uploaded files.&lt;/li&gt;
&lt;/ul&gt;

&lt;h3&gt;
  
  
  4. Set up live database replication
&lt;/h3&gt;

&lt;ul&gt;
&lt;li&gt;Rather than taking your app offline for hours to run &lt;code&gt;pg_dump&lt;/code&gt; and &lt;code&gt;pg_restore&lt;/code&gt;, use logical replication (for PostgreSQL) or binary log replication (for MySQL).&lt;/li&gt;
&lt;li&gt;Configure your new EU database to subscribe to changes from your DigitalOcean Managed Database over an encrypted connection, restricting access by IP address in your DigitalOcean database firewall.&lt;/li&gt;
&lt;li&gt;Let the initial data copy complete, and verify that replication lag drops to near zero. Your EU database is now a live, up-to-date follower of your DigitalOcean database.&lt;/li&gt;
&lt;/ul&gt;

&lt;h3&gt;
  
  
  5. Point your CI/CD pipelines and test
&lt;/h3&gt;

&lt;ul&gt;
&lt;li&gt;Update your GitHub Actions, GitLab CI, or deployment scripts to deploy your application containers or binaries to the new EU servers or Kubernetes cluster.&lt;/li&gt;
&lt;li&gt;Update your application's environment variables in the new environment so it points to the new EU database, the new Redis cache, and the new S3 object storage endpoints.&lt;/li&gt;
&lt;li&gt;Test the new environment thoroughly using internal host file overrides or a staging domain before exposing it to public traffic.&lt;/li&gt;
&lt;/ul&gt;

&lt;h3&gt;
  
  
  6. Cut over traffic and keep a rollback path
&lt;/h3&gt;

&lt;ul&gt;
&lt;li&gt;Choose a low-traffic window for your users.&lt;/li&gt;
&lt;li&gt;Put your application into a brief maintenance or read-only mode (so no new writes hit the DigitalOcean database).&lt;/li&gt;
&lt;li&gt;Verify that database replication lag is zero and run a final fast &lt;code&gt;rclone&lt;/code&gt; pass on your object storage.&lt;/li&gt;
&lt;li&gt;Promote the EU database to be the primary writer.&lt;/li&gt;
&lt;li&gt;Update your DNS records (or proxy targets in Cloudflare/bunny.net) to point to the new EU Load Balancer IP address.&lt;/li&gt;
&lt;li&gt;Keep the old DigitalOcean environment running for a week or two (ideally with reverse replication configured if you need a fast rollback option) before deleting the old resources and closing out your bill.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Common pitfalls to watch out for
&lt;/h2&gt;

&lt;p&gt;Even in a simple migration, a few details can catch teams by surprise:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Hardcoded Spaces URLs:&lt;/strong&gt; DigitalOcean Spaces URLs follow the pattern &lt;code&gt;bucket.region.digitaloceanspaces.com&lt;/code&gt; (or &lt;code&gt;cdn.digitaloceanspaces.com&lt;/code&gt;). If you stored full URLs in your database rows instead of relative paths, you will need to run a database migration script to update those strings, or use a custom domain for your object storage so the URL never changes again.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Droplet metadata service:&lt;/strong&gt; If your custom boot scripts query the DigitalOcean metadata IP (&lt;code&gt;169.254.169.254&lt;/code&gt;) using DigitalOcean-specific JSON paths, you may need to adjust them to standard &lt;code&gt;cloud-init&lt;/code&gt; formats used by UpCloud, Scaleway, and Hetzner.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Database connection pools:&lt;/strong&gt; DigitalOcean Managed Databases include built-in PgBouncer connection pooling configured in the control panel. Check how your target provider handles connection pooling (UpCloud and Scaleway offer managed pooling options, while on Hetzner you will need to install PgBouncer alongside PostgreSQL).&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;SMTP port blocking:&lt;/strong&gt; To prevent spam, DigitalOcean, UpCloud, Scaleway, and Hetzner all frequently block outbound ports 25, 465, and 587 by default on new accounts. If your servers send email directly or connect to an SMTP relay over those ports, request port unblocking from support &lt;em&gt;before&lt;/em&gt; your migration night, or switch to an HTTP API or port 2525.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Frequently asked questions
&lt;/h2&gt;

&lt;h3&gt;
  
  
  Will moving from DigitalOcean to an EU cloud require downtime?
&lt;/h3&gt;

&lt;p&gt;Most migrations require a short, planned write pause during the final cutover. Because your database and object storage are synchronized in the background while your app stays live, the actual cutover only requires pausing database writes, confirming replication lag is under a few seconds, promoting the new database, and switching DNS. For most small teams, this write pause takes between a few seconds and a few minutes.&lt;/p&gt;

&lt;h3&gt;
  
  
  How do we migrate DigitalOcean App Platform?
&lt;/h3&gt;

&lt;p&gt;DigitalOcean App Platform is a Platform-as-a-Service (PaaS) that builds and runs code from your Git repository. If your app uses a Dockerfile, the easiest managed equivalent in Europe is Scaleway Serverless Containers. Alternatively, if you move to UpCloud or Hetzner, you can deploy your containers onto a Managed Kubernetes cluster (UpCloud) or use a lightweight deployment tool like Kamal, Docker Compose, or Coolify on a virtual machine.&lt;/p&gt;

&lt;h3&gt;
  
  
  Can we keep Cloudflare in front of our new EU infrastructure?
&lt;/h3&gt;

&lt;p&gt;Yes, and doing so actually makes the cutover faster. If you proxy your traffic through Cloudflare (or a European alternative like bunny.net), you do not have to wait for global DNS caches to expire during the switch. When you change the origin IP address in your proxy dashboard from your DigitalOcean Load Balancer to your new EU Load Balancer, traffic shifts to the new environment almost immediately.&lt;/p&gt;

&lt;h3&gt;
  
  
  Which provider is best if we need GDPR and DORA compliance?
&lt;/h3&gt;

&lt;p&gt;UpCloud, Scaleway, and Hetzner are all headquartered in the European Union (Finland, France, and Germany respectively) and operate EU data centers under European jurisdiction, avoiding US extraterritorial laws that apply to US-owned providers like DigitalOcean. If your company falls under the EU Digital Operational Resilience Act (DORA, which applied from January 17, 2025), moving to standard open-source infrastructure defined in Terraform also satisfies key requirements for documented ICT exit strategies and reduced vendor lock-in.&lt;/p&gt;

&lt;h2&gt;
  
  
  Sources
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;&lt;a href="https://eur-lex.europa.eu/eli/reg/2023/2854/oj" rel="noopener noreferrer"&gt;Regulation (EU) 2023/2854 (EU Data Act) on EUR-Lex&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://eur-lex.europa.eu/eli/reg/2022/2554/oj" rel="noopener noreferrer"&gt;Regulation (EU) 2022/2554 (Digital Operational Resilience Act / DORA) on EUR-Lex&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://www.digitalocean.com/pricing" rel="noopener noreferrer"&gt;DigitalOcean Pricing Overview&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://upcloud.com/pricing/" rel="noopener noreferrer"&gt;UpCloud Pricing and Plans&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://www.scaleway.com/en/pricing/" rel="noopener noreferrer"&gt;Scaleway Cloud Pricing&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://www.hetzner.com/cloud" rel="noopener noreferrer"&gt;Hetzner Cloud Overview and Pricing&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;




&lt;p&gt;&lt;em&gt;Drafted with AI assistance and fact-checked by the MoveInfra team. The original, kept up to date, is at &lt;a href="https://moveinfra.eu/blog/digitalocean-to-eu-cloud/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu&lt;/a&gt;. Free tools for pricing a move to an EU cloud: &lt;a href="https://moveinfra.eu/tools/?utm_source=devto&amp;amp;utm_medium=syndication" rel="noopener noreferrer"&gt;moveinfra.eu/tools&lt;/a&gt;.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>cloud</category>
      <category>digitalocean</category>
      <category>upcloud</category>
      <category>scaleway</category>
    </item>
  </channel>
</rss>
