<?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: Doğukan Sağlık</title>
    <description>The latest articles on DEV Community by Doğukan Sağlık (@lazypanic999).</description>
    <link>https://dev.to/lazypanic999</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%2F4213951%2Fac069784-077b-4399-9f3a-0bc947e29680.jpg</url>
      <title>DEV Community: Doğukan Sağlık</title>
      <link>https://dev.to/lazypanic999</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/lazypanic999"/>
    <language>en</language>
    <item>
      <title>Restivra: verify PostgreSQL backups with real restore tests</title>
      <dc:creator>Doğukan Sağlık</dc:creator>
      <pubDate>Fri, 20 Nov 2026 00:52:58 +0000</pubDate>
      <link>https://dev.to/lazypanic999/restivra-restoring-postgresql-backups-in-disposable-containers-12pl</link>
      <guid>https://dev.to/lazypanic999/restivra-restoring-postgresql-backups-in-disposable-containers-12pl</guid>
      <description>&lt;p&gt;Restivra restores an existing PostgreSQL backup into a disposable container, runs your SQL checks against the restored data, and keeps the result in a local web dashboard. It is free, open source under MIT, and runs with Docker Compose.&lt;/p&gt;

&lt;h2&gt;
  
  
  Watch the 51-second walkthrough
&lt;/h2&gt;

&lt;p&gt;&lt;a href="https://lazypanic999.github.io/restivra/" rel="noopener noreferrer"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F55pjamh68jyt94beqmgd.jpg" alt="Watch Restivra: create a project, restore a backup, inspect a report and catch stale data" width="800" height="450"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Click the preview to watch the video in your browser. It has sound and uses synthetic data. The current pilot also adds retained-history filters and CSV exports.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://github.com/lazypanic999/restivra" rel="noopener noreferrer"&gt;View the project on GitHub&lt;/a&gt; · &lt;a href="https://github.com/lazypanic999/restivra/releases/tag/v0.1.0-pilot.1" rel="noopener noreferrer"&gt;Try the pinned pilot&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Below is the restore workflow, two recovery edge cases from testing, and a runnable setup.&lt;/p&gt;

&lt;h2&gt;
  
  
  From an archive to a result
&lt;/h2&gt;

&lt;p&gt;Restivra starts with a completed PostgreSQL 17 custom-format archive produced by pg_dump 17 with -Fc. It does not connect to your production database to create a backup.&lt;/p&gt;

&lt;p&gt;A Go worker takes a private snapshot of the archive, calculates its SHA-256, and restores it into a temporary PostgreSQL container. The restore container has no external network or published ports. It runs under resource limits, and the worker removes it after verification.&lt;/p&gt;

&lt;p&gt;You can add application-specific SQL checks. Each check must return exactly one non-null boolean. For example, if the application should have orders:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight sql"&gt;&lt;code&gt;&lt;span class="k"&gt;SELECT&lt;/span&gt; &lt;span class="nb"&gt;count&lt;/span&gt;&lt;span class="cm"&gt;(*) &amp;gt; 0 FROM public.orders
&lt;/span&gt;&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;That check would be wrong for a legitimately empty application. The checks need to express your own invariants. A successful restore with no meaningful checks provides less evidence about whether the data is useful.&lt;/p&gt;

&lt;p&gt;Checks run in read-only transactions using a restricted role, with a ten-second statement timeout. The UI records restore and check outcomes, cleanup, archive fingerprint and attempt history. It defaults to English and also supports Turkish.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fjj3bx7occd0207jhun3a.jpg" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fjj3bx7occd0207jhun3a.jpg" alt="Restivra retained run history with synthetic projects" width="800" height="721"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  Two recovery details worth testing
&lt;/h2&gt;

&lt;p&gt;The queue and attempt history live in PostgreSQL. A database advisory lock restricts an installation to one restore worker at a time.&lt;/p&gt;

&lt;p&gt;One cleanup edge case is a Docker create operation that finishes after its caller has stopped waiting. An initial absence check can succeed before that late container appears. Startup cleanup alone would leave it around until another restart. The worker now reconciles installation-owned resources before the next restore and roughly every 30 idle seconds. Unknown cleanup stops new work; another installation's resources must be preserved.&lt;/p&gt;

&lt;p&gt;Another edge case is a delayed completion from an older worker attempt. It must not overwrite a newer running attempt. Completion checks the expected attempt number under the job's row lock. Earlier failures remain in attempt history even when a fresh retry passes.&lt;/p&gt;

&lt;p&gt;Transient infrastructure failures have bounded retries. A failed application check or invalid archive stays a failure. Retrying those automatically would not make the underlying evidence better.&lt;/p&gt;

&lt;h2&gt;
  
  
  Try the pinned pilot
&lt;/h2&gt;

&lt;p&gt;The primary handoff is Windows with Git, PowerShell 7.2+ and Docker Desktop using Linux containers. Go, Node and a host PostgreSQL installation are not required; the application builds inside Docker.&lt;/p&gt;

&lt;p&gt;Start Docker, then run in PowerShell:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight powershell"&gt;&lt;code&gt;&lt;span class="n"&gt;git&lt;/span&gt; &lt;span class="n"&gt;clone&lt;/span&gt; &lt;span class="o"&gt;--&lt;/span&gt;&lt;span class="n"&gt;branch&lt;/span&gt; &lt;span class="n"&gt;v0&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="mf"&gt;1.0&lt;/span&gt;&lt;span class="o"&gt;-&lt;/span&gt;&lt;span class="n"&gt;pilot&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="mi"&gt;1&lt;/span&gt; &lt;span class="o"&gt;--&lt;/span&gt;&lt;span class="n"&gt;depth&lt;/span&gt; &lt;span class="mi"&gt;1&lt;/span&gt; &lt;span class="n"&gt;https&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="o"&gt;//&lt;/span&gt;&lt;span class="n"&gt;github&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="n"&gt;com&lt;/span&gt;&lt;span class="o"&gt;/&lt;/span&gt;&lt;span class="n"&gt;lazypanic999&lt;/span&gt;&lt;span class="o"&gt;/&lt;/span&gt;&lt;span class="n"&gt;restivra&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="n"&gt;git&lt;/span&gt; &lt;span class="n"&gt;restivra&lt;/span&gt;&lt;span class="o"&gt;-&lt;/span&gt;&lt;span class="n"&gt;pilot&lt;/span&gt;
&lt;span class="n"&gt;cd&lt;/span&gt; &lt;span class="n"&gt;restivra&lt;/span&gt;&lt;span class="o"&gt;-&lt;/span&gt;&lt;span class="n"&gt;pilot&lt;/span&gt;
&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="o"&gt;/&lt;/span&gt;&lt;span class="n"&gt;scripts&lt;/span&gt;&lt;span class="o"&gt;/&lt;/span&gt;&lt;span class="n"&gt;setup&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="n"&gt;ps1&lt;/span&gt;
&lt;span class="n"&gt;docker&lt;/span&gt; &lt;span class="n"&gt;pull&lt;/span&gt; &lt;span class="n"&gt;postgres&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="mi"&gt;17&lt;/span&gt;&lt;span class="o"&gt;-&lt;/span&gt;&lt;span class="n"&gt;bookworm&lt;/span&gt;
&lt;span class="n"&gt;docker&lt;/span&gt; &lt;span class="n"&gt;compose&lt;/span&gt; &lt;span class="n"&gt;up&lt;/span&gt; &lt;span class="o"&gt;--&lt;/span&gt;&lt;span class="n"&gt;build&lt;/span&gt; &lt;span class="o"&gt;-&lt;/span&gt;&lt;span class="n"&gt;d&lt;/span&gt; &lt;span class="o"&gt;--&lt;/span&gt;&lt;span class="n"&gt;wait&lt;/span&gt; &lt;span class="o"&gt;--&lt;/span&gt;&lt;span class="n"&gt;wait&lt;/span&gt;&lt;span class="o"&gt;-&lt;/span&gt;&lt;span class="n"&gt;timeout&lt;/span&gt; &lt;span class="mi"&gt;180&lt;/span&gt;
&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="o"&gt;/&lt;/span&gt;&lt;span class="n"&gt;scripts&lt;/span&gt;&lt;span class="o"&gt;/&lt;/span&gt;&lt;span class="n"&gt;pilot&lt;/span&gt;&lt;span class="o"&gt;-&lt;/span&gt;&lt;span class="n"&gt;check&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="n"&gt;ps1&lt;/span&gt;
&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="o"&gt;/&lt;/span&gt;&lt;span class="n"&gt;scripts&lt;/span&gt;&lt;span class="o"&gt;/&lt;/span&gt;&lt;span class="n"&gt;pilot&lt;/span&gt;&lt;span class="o"&gt;-&lt;/span&gt;&lt;span class="n"&gt;smoke&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="n"&gt;ps1&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Open &lt;a href="http://127.0.0.1:7474" rel="noopener noreferrer"&gt;http://127.0.0.1:7474&lt;/a&gt; and sign in with RESTIVRA_AUTH_TOKEN from your private .env file. Do not share that file.&lt;/p&gt;

&lt;p&gt;The smoke script creates a real PostgreSQL archive containing two synthetic customers. It then creates two Pilot projects: one checks the expected customer count; the other intentionally runs SELECT false. The second project should fail its SQL check while restore and cleanup succeed. Schedules stay disabled until you configure them.&lt;/p&gt;

&lt;p&gt;After inspecting those reports, you can archive the sample projects and add your own authorized archive. Daily schedules, history filters and CSV exports are included. A sleeping computer cannot run scheduled checks; after returning, an overdue daily schedule queues one run rather than a backlog for every missed day.&lt;/p&gt;

&lt;h2&gt;
  
  
  Current limits and feedback
&lt;/h2&gt;

&lt;p&gt;This is an early localhost pilot for trusted PostgreSQL 17 archives. Defaults are a 256 MiB archive, 512 MiB RAM-backed restored data, and 1536 MiB restore-container memory. A small compressed archive can expand past the data limit.&lt;/p&gt;

&lt;p&gt;Custom extensions, other major versions, PITR, public uploads and hosted multi-tenancy are outside the validated profile. Containers are not a hostile-dump sandbox. A pass also does not prove original permissions, full application recovery or backup freshness. Keep your existing backup and independent recovery procedure.&lt;/p&gt;

&lt;p&gt;Fresh setup and the synthetic restore checks passed locally on Windows and in Linux CI using PowerShell. The external seven-day developer pilot is still pending. Engineering tests are not a substitute for that pilot.&lt;/p&gt;

&lt;p&gt;You can start with the synthetic demo before using any real archive. Feedback on installation, report clarity and useful SQL checks is welcome. No real dumps or credentials are needed for feedback.&lt;/p&gt;

&lt;p&gt;If Restivra is useful to you, consider starring the repository to follow its development or sharing it with someone who maintains PostgreSQL backups.&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;&lt;a href="https://github.com/lazypanic999/restivra" rel="noopener noreferrer"&gt;Source and MIT license&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://github.com/lazypanic999/restivra/releases/tag/v0.1.0-pilot.1" rel="noopener noreferrer"&gt;Pinned pilot and English/Turkish setup guides&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://github.com/lazypanic999/restivra/tree/v0.1.0-pilot.1/docs/acceptance/pilot-handoff-20261120" rel="noopener noreferrer"&gt;Engineering acceptance evidence&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;AI disclosure: substantial implementation, tests, documentation and this post were created with OpenAI Codex. Automated tests have been run; independent review and real-user feedback remain necessary.&lt;/p&gt;

</description>
      <category>postgres</category>
      <category>devops</category>
      <category>opensource</category>
      <category>docker</category>
    </item>
    <item>
      <title>A small ACK can hide a failed return path: collecting both endpoint logs</title>
      <dc:creator>Doğukan Sağlık</dc:creator>
      <pubDate>Sun, 08 Nov 2026 18:03:42 +0000</pubDate>
      <link>https://dev.to/lazypanic999/a-small-ack-can-hide-a-failed-return-path-collecting-both-endpoint-logs-9d7</link>
      <guid>https://dev.to/lazypanic999/a-small-ack-can-hide-a-failed-return-path-collecting-both-endpoint-logs-9d7</guid>
      <description>&lt;p&gt;When a tunnel or routed path passes &lt;code&gt;ping&lt;/code&gt; but stalls on bulk transfers, testing large UDP payloads in one direction is only half the story. If a receiver answers a 1400-byte probe with a 24-byte acknowledgement, an &lt;code&gt;acknowledged&lt;/code&gt; result only records that 1400 bytes arrived forward and 24 bytes made it back. When the return path has a narrower MTU, small ACKs can succeed while large return payloads fail.&lt;/p&gt;

&lt;p&gt;I maintain &lt;a href="https://github.com/lazypanic999/tunnelscope" rel="noopener noreferrer"&gt;TunnelScope&lt;/a&gt;, an alpha Python CLI that combines bounded UDP probes, &lt;code&gt;--sized-reply&lt;/code&gt;, and offline endpoint log correlation. With &lt;code&gt;--sized-reply&lt;/code&gt; enabled on both sides, a Linux (including WSL2) or Windows receiver attempts a reply of the probe's own payload size. Joining both logs records receiver arrival, the receiver socket's send outcome, and whether the sender received that reply. A missing receiver entry is inconclusive without verified capture coverage; a successful socket send alone does not locate a later failure. A timeout alone does not establish an MTU fault.&lt;/p&gt;

&lt;h2&gt;
  
  
  What the controlled asymmetric-return lab shows
&lt;/h2&gt;

&lt;p&gt;In the recorded &lt;a href="https://github.com/lazypanic999/tunnelscope/blob/main/docs/validation/asymmetric-return.md" rel="noopener noreferrer"&gt;asymmetric return-path lab&lt;/a&gt;, forward traffic crosses 1500-byte links from sender A through router R to receiver B, while return traffic from R to A crosses a 1280-byte link.&lt;/p&gt;

&lt;p&gt;With &lt;code&gt;--dont-fragment --sized-reply&lt;/code&gt;, 64- and 1200-byte payloads arrive and their same-sized replies return. Both 1400-byte probes arrive, but their replies do not come back. The recorded first attempt shows a receiver socket send accepted locally and a missing reply at the sender. After ICMP feedback, the recorded second attempt shows a local &lt;code&gt;EMSGSIZE&lt;/code&gt; send failure. An ordinary 1400-byte probe is still acknowledged because its 24-byte ACK fits the return link.&lt;/p&gt;

&lt;p&gt;This topology establishes the constraint in that lab. On an unknown path, the logs alone do not locate a missing reply.&lt;/p&gt;

&lt;h2&gt;
  
  
  Try the workflow locally on loopback
&lt;/h2&gt;

&lt;p&gt;The example below uses a Linux/WSL2 shell. Start in a fresh directory; this checks the workflow, not a remote VPN:&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;mkdir &lt;/span&gt;tunnelscope-demo &lt;span class="o"&gt;&amp;amp;&amp;amp;&lt;/span&gt; &lt;span class="nb"&gt;cd &lt;/span&gt;tunnelscope-demo
python3 &lt;span class="nt"&gt;-m&lt;/span&gt; venv .venv
&lt;span class="nb"&gt;source&lt;/span&gt; .venv/bin/activate
python &lt;span class="nt"&gt;-m&lt;/span&gt; pip &lt;span class="nb"&gt;install &lt;/span&gt;&lt;span class="nv"&gt;tunnelscope&lt;/span&gt;&lt;span class="o"&gt;==&lt;/span&gt;0.9.5

&lt;span class="c"&gt;# Terminal 1: bounded loopback receiver&lt;/span&gt;
tunnelscope receive &lt;span class="nt"&gt;--max-packets&lt;/span&gt; 2 &lt;span class="nt"&gt;--idle-timeout&lt;/span&gt; 30 &lt;span class="nt"&gt;--sized-reply&lt;/span&gt; &lt;span class="nt"&gt;--output&lt;/span&gt; receiver.json

&lt;span class="c"&gt;# Terminal 2: activate the same venv in this shell&lt;/span&gt;
&lt;span class="nb"&gt;source&lt;/span&gt; .venv/bin/activate
tunnelscope probe 127.0.0.1 &lt;span class="nt"&gt;--sizes&lt;/span&gt; 64,1400 &lt;span class="nt"&gt;--count&lt;/span&gt; 1 &lt;span class="nt"&gt;--sized-reply&lt;/span&gt; &lt;span class="nt"&gt;--output&lt;/span&gt; probe.json
tunnelscope correlate probe.json receiver.json &lt;span class="nt"&gt;--output&lt;/span&gt; correlation.json &lt;span class="nt"&gt;--html&lt;/span&gt; correlation.html
&lt;span class="nb"&gt;mkdir &lt;/span&gt;shared
tunnelscope sanitize probe.json receiver.json correlation.json &lt;span class="nt"&gt;--output-dir&lt;/span&gt; shared &lt;span class="nt"&gt;--html&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Wait for &lt;code&gt;receive&lt;/code&gt; to finish writing &lt;code&gt;receiver.json&lt;/code&gt; before running &lt;code&gt;correlate&lt;/code&gt;. Sanitize the sender, receiver and correlation files together so replacements stay consistent. Native Windows users can follow the &lt;a href="https://github.com/lazypanic999/tunnelscope/blob/main/docs/windows-loopback.md" rel="noopener noreferrer"&gt;PowerShell loopback guide&lt;/a&gt;; Linux-only DF/ICMP options remain separate.&lt;/p&gt;

&lt;h2&gt;
  
  
  Scope and what v0.9.5 changes
&lt;/h2&gt;

&lt;p&gt;Same-sized replies, correlation and sanitization predate v0.9.5. Existing lab and VM records keep their original producer versions. Current evidence is from controlled namespace, loopback and single-host VM environments, not independent production validation. The v0.9.5 patch hardens offline input validation, IPv6 and placeholder collision handling, and exclusive output-file creation. It changes no wire protocol, JSON schema or measurement logic.&lt;/p&gt;

&lt;p&gt;Sanitization is best effort. Inspect the copies and filenames before sharing. The UDP receiver is unauthenticated: use controlled endpoints, bind to the intended address and restrict access to the sender.&lt;/p&gt;

&lt;h2&gt;
  
  
  One operator feedback request
&lt;/h2&gt;

&lt;p&gt;Would the &lt;a href="https://github.com/lazypanic999/tunnelscope/blob/main/docs/field-reports.md" rel="noopener noreferrer"&gt;field report guide&lt;/a&gt; be practical on a two-host WireGuard, OpenVPN or IPsec path you are allowed to test? A reviewed, sanitized report from a healthy path is useful too. &lt;a href="https://github.com/lazypanic999/tunnelscope/issues/2" rel="noopener noreferrer"&gt;Issue #2&lt;/a&gt; tracks independent VPN evidence. No credentials, private captures or remote access are needed.&lt;/p&gt;

</description>
      <category>networking</category>
      <category>linux</category>
      <category>python</category>
      <category>opensource</category>
    </item>
    <item>
      <title>A ready 5G pod can still drop packets: CorePulse needs contributors</title>
      <dc:creator>Doğukan Sağlık</dc:creator>
      <pubDate>Tue, 03 Nov 2026 22:47:49 +0000</pubDate>
      <link>https://dev.to/lazypanic999/a-ready-5g-pod-can-still-drop-packets-corepulse-needs-contributors-30f2</link>
      <guid>https://dev.to/lazypanic999/a-ready-5g-pod-can-still-drop-packets-corepulse-needs-contributors-30f2</guid>
      <description>&lt;p&gt;I’m building &lt;a href="https://github.com/lazypanic999/CorePulse" rel="noopener noreferrer"&gt;CorePulse&lt;/a&gt;, an MIT-licensed Go project for controlled fault experiments in a local Open5GS + UERANSIM Kubernetes lab. I’ve opened six scoped issues for people who want to help with Go, documentation, or telecom observability.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fybtcmpi0474f42ae3opv.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fybtcmpi0474f42ae3opv.png" alt="CorePulse dashboard showing the local lab" width="800" height="967"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  The result that changed the recovery checks
&lt;/h2&gt;

&lt;p&gt;In a recent isolated UPF pod replacement experiment, Kubernetes made the replacement ready, but six subsequent UE-to-UPF gateway checks still lost both packets. Resource telemetry and the experiment interval were recorded. A green deployment badge was only one part of the recovery result.&lt;/p&gt;

&lt;p&gt;The reset procedure now stops UE, gNB, SMF and UPF, waits for the old pods to disappear, then starts UPF, SMF, gNB and UE in order. In the pinned lab, routing PFCP through a separate headless service let SMF resolve the current UPF pod directly. After the reset, UE registration and PDU session establishment returned, two consecutive gateway checks passed, and the unchanged lab verifier passed all 16 deployments and 5/5 external ping replies.&lt;/p&gt;

&lt;p&gt;This is a local result for the tested chart and image versions. It does not establish a general root cause for Kubernetes PFCP failures. The &lt;a href="https://github.com/lazypanic999/CorePulse/blob/main/docs/lab-integration.md" rel="noopener noreferrer"&gt;integration notes&lt;/a&gt; include the failed attempts and limits.&lt;/p&gt;

&lt;h2&gt;
  
  
  What works today
&lt;/h2&gt;

&lt;p&gt;The Go CLI and local dashboard support guarded AMF/UPF pod deletion, real Kubernetes CPU/memory collection, explicit normal-run collection, run records and repeated packet-path checks. Curated CSV exports include only qualifying run intervals; unscoped or unverified samples are excluded.&lt;/p&gt;

&lt;p&gt;There is no trained detector yet. The current curated dataset has 12 normal and 12 fault rows across five qualifying runs, which is much too small for a meaningful model evaluation. Registration, session and PFCP log events still need a structured feature pipeline.&lt;/p&gt;

&lt;h2&gt;
  
  
  Where you can help
&lt;/h2&gt;

&lt;p&gt;Two tasks are labeled &lt;strong&gt;good first issue&lt;/strong&gt; and need no Kubernetes cluster:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;a href="https://github.com/lazypanic999/CorePulse/issues/1" rel="noopener noreferrer"&gt;Reject negative and non-finite resource quantities (#1)&lt;/a&gt;: a focused Go validation change with fixture tests.&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://github.com/lazypanic999/CorePulse/issues/2" rel="noopener noreferrer"&gt;Write a complete offline contributor quickstart (#2)&lt;/a&gt;: help someone build, test and preview the dashboard before setting up a lab.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Four &lt;strong&gt;help wanted&lt;/strong&gt; tasks go deeper:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;a href="https://github.com/lazypanic999/CorePulse/issues/3" rel="noopener noreferrer"&gt;Persist fault checkpoints before recovery finishes (#3)&lt;/a&gt;.&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://github.com/lazypanic999/CorePulse/issues/4" rel="noopener noreferrer"&gt;Make baseline collection cancellable (#4)&lt;/a&gt;.&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://github.com/lazypanic999/CorePulse/issues/5" rel="noopener noreferrer"&gt;Distinguish missing metrics from measured zeros in CSV (#5)&lt;/a&gt;.&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://github.com/lazypanic999/CorePulse/issues/6" rel="noopener noreferrer"&gt;Parse timestamped registration, session and PFCP events (#6)&lt;/a&gt;.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Each issue includes starting files, acceptance criteria and validation expectations. Comment on the one you want to tackle with a short approach so we can coordinate. The &lt;a href="https://github.com/lazypanic999/CorePulse/blob/main/CONTRIBUTING.md" rel="noopener noreferrer"&gt;contribution guide&lt;/a&gt; covers checks and lab boundaries.&lt;/p&gt;

&lt;p&gt;Development and this write-up have been assisted by AI tools. The lab observations above come from actual runs; the project’s limitations remain documented.&lt;/p&gt;

&lt;p&gt;If you work with Open5GS or UERANSIM, I’d also welcome a review of the recovery procedure and which protocol events would be most useful to collect next.&lt;/p&gt;

</description>
      <category>opensource</category>
      <category>go</category>
      <category>kubernetes</category>
      <category>devops</category>
    </item>
    <item>
      <title>I built TunnelScope to inspect MTU and packet-size failures with evidence</title>
      <dc:creator>Doğukan Sağlık</dc:creator>
      <pubDate>Sun, 01 Nov 2026 18:15:30 +0000</pubDate>
      <link>https://dev.to/lazypanic999/i-built-tunnelscope-to-inspect-mtu-and-packet-size-failures-with-evidence-b4p</link>
      <guid>https://dev.to/lazypanic999/i-built-tunnelscope-to-inspect-mtu-and-packet-size-failures-with-evidence-b4p</guid>
      <description>&lt;p&gt;Small pings can work while larger traffic stalls. A sender timeout leaves several possibilities open: the probe failed to arrive, its ACK failed to return, or something else interrupted the exchange. I built &lt;strong&gt;TunnelScope&lt;/strong&gt;, an MIT-licensed Python CLI, to keep the observations needed to investigate that uncertainty on a controlled path.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Update — 8 November 2026:&lt;/strong&gt; &lt;a href="https://pypi.org/project/tunnelscope/0.9.3/" rel="noopener noreferrer"&gt;v0.9.3 alpha is now on PyPI&lt;/a&gt;. The v0.9 line adds bounded Markdown summaries, declared report/input versions and controlled OpenVPN return-path evidence. The latest patch rejects non-probe sender files before correlation outputs are written, preserves missing-kind legacy reports, and records nonzero local IPv6 receiver scopes. Python 3.13 delivery coverage and tests run from the built source archive remain in place. Missing receiver send outcomes stay explicitly unknown. &lt;a href="https://github.com/lazypanic999/tunnelscope/releases/tag/v0.9.3" rel="noopener noreferrer"&gt;Release notes&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fk53dpv1hc0dd7yc63fez.gif" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fk53dpv1hc0dd7yc63fez.gif" alt="30-second replay of recorded OpenVPN return-path evidence and a sanitized Markdown summary" width="800" height="475"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;This 30-second demo shows selected observations from recorded OpenVPN evidence and reruns the correlation, batch sanitization and Markdown steps. It is not a live network capture. The &lt;a href="https://github.com/lazypanic999/tunnelscope/blob/main/docs/demo-v09.md" rel="noopener noreferrer"&gt;v0.9 replay guide and input reports&lt;/a&gt; explain how to reproduce the report processing; the measurement input and report-processing versions are shown separately.&lt;/p&gt;

&lt;h2&gt;
  
  
  Start by joining both endpoint logs
&lt;/h2&gt;

&lt;p&gt;In an empty directory on Linux or WSL2:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;python3 &lt;span class="nt"&gt;-m&lt;/span&gt; venv .venv
&lt;span class="nb"&gt;source&lt;/span&gt; .venv/bin/activate
python &lt;span class="nt"&gt;-m&lt;/span&gt; pip &lt;span class="nb"&gt;install &lt;/span&gt;&lt;span class="nv"&gt;tunnelscope&lt;/span&gt;&lt;span class="o"&gt;==&lt;/span&gt;0.9.3
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;In terminal 1, start a loopback receiver:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;tunnelscope receive &lt;span class="nt"&gt;--max-packets&lt;/span&gt; 2 &lt;span class="nt"&gt;--output&lt;/span&gt; receiver.json
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Activate the same environment in terminal 2. After the receiver says it is listening, run:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;tunnelscope probe 127.0.0.1 &lt;span class="nt"&gt;--sizes&lt;/span&gt; 64,1200 &lt;span class="nt"&gt;--count&lt;/span&gt; 1 &lt;span class="nt"&gt;--output&lt;/span&gt; probe.json
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Once the receiver exits and writes its capture:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;tunnelscope correlate probe.json receiver.json &lt;span class="nt"&gt;--output&lt;/span&gt; correlation.json &lt;span class="nt"&gt;--html&lt;/span&gt; correlation.html
tunnelscope show correlation.json
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Open &lt;code&gt;correlation.html&lt;/code&gt; to see both endpoints' observations by probe ID. This quick start uses small ACKs. For same-sized return evidence, start both endpoints with &lt;code&gt;--sized-reply&lt;/code&gt;; the &lt;a href="https://github.com/lazypanic999/tunnelscope/blob/main/docs/sized-reply.md" rel="noopener noreferrer"&gt;sized-reply guide&lt;/a&gt; explains the observations. This loopback example needs no root privileges or network setup. For native PowerShell, the &lt;a href="https://github.com/lazypanic999/tunnelscope/blob/main/docs/windows-loopback.md" rel="noopener noreferrer"&gt;Windows loopback guide&lt;/a&gt; covers ordinary UDP probes and report processing; DF and ICMP observation options require Linux support.&lt;/p&gt;

&lt;h2&gt;
  
  
  New in v0.9: share a reviewable evidence summary
&lt;/h2&gt;

&lt;p&gt;After joining the quick-start reports, sanitize the files together and export a bounded issue summary:&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;mkdir &lt;/span&gt;shared
tunnelscope sanitize probe.json receiver.json correlation.json &lt;span class="nt"&gt;--output-dir&lt;/span&gt; shared
tunnelscope show shared/correlation.json &lt;span class="nt"&gt;--format&lt;/span&gt; markdown &lt;span class="o"&gt;&amp;gt;&lt;/span&gt; evidence.md
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Review the copies before sharing. The Markdown exporter does not sanitize its input; it keeps at most 50 observation rows and 24,000 ASCII characters, explicitly notes truncation, and escapes report values as text. Attach the full sanitized JSON alongside the summary. Producer/input versions are declared metadata, and missing legacy versions remain unknown. &lt;a href="https://github.com/lazypanic999/tunnelscope/blob/main/docs/markdown-reports.md" rel="noopener noreferrer"&gt;Export guide&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  What changed in v0.6
&lt;/h2&gt;

&lt;p&gt;&lt;code&gt;diagnose&lt;/code&gt; runs three bounded, sequential experiments from one Linux sender: DF probes, the same sizes with source fragmentation permitted, and adaptive payload-size search. It writes JSON and offline HTML while retaining all underlying attempts and run IDs. A cooperating receiver must already be running:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;tunnelscope diagnose PEER_IP &lt;span class="nt"&gt;--min&lt;/span&gt; 1200 &lt;span class="nt"&gt;--max&lt;/span&gt; 1400 &lt;span class="nt"&gt;--count&lt;/span&gt; 2 &lt;span class="nt"&gt;--output&lt;/span&gt; diagnosis.json
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;After collecting the receiver's completed capture, the same &lt;code&gt;correlate&lt;/code&gt; command accepts the diagnosis report and joins all three runs:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;tunnelscope correlate diagnosis.json receiver.json &lt;span class="nt"&gt;--output&lt;/span&gt; joined.json &lt;span class="nt"&gt;--html&lt;/span&gt; joined.html
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Feedback on the original post asked for a joined view that distinguishes “receiver observed the probe, sender got no ACK” from “no ACK and no receiver observation.” The report now keeps those cases separate, alongside payload size, direction and per-attempt timestamps. Missing receiver entries remain inconclusive without verified capture coverage; clocks on different hosts are shown without comparing them.&lt;/p&gt;

&lt;p&gt;Repeated measurements also include a descriptive no-ACK fraction and Wilson interval. Linux UDP error-queue observations distinguish local size errors from ICMP fragmentation-needed or IPv6 Packet Too Big feedback. Neither timeouts nor those intervals are a measured network packet-loss rate.&lt;/p&gt;

&lt;h2&gt;
  
  
  What I have checked
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Two WireGuard VMs:&lt;/strong&gt; independent Ubuntu and Debian guests exchanged the same payload sizes in both directions. The observed 1252/1253-byte UDP boundary reflected their configured local 1280-byte tunnel interface MTU. &lt;a href="https://github.com/lazypanic999/tunnelscope/blob/main/docs/validation/two-vm-wireguard.md" rel="noopener noreferrer"&gt;Validation record&lt;/a&gt;.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Routed WireGuard constraint:&lt;/strong&gt; a receiving gateway forwarded onto a lower-MTU LAN. Both directions recorded matching ICMP when feedback was visible. Blocking that feedback produced timeouts and an inconclusive search. This was a four-namespace lab on one kernel. &lt;a href="https://github.com/lazypanic999/tunnelscope/blob/main/docs/validation/wireguard-routed.md" rel="noopener noreferrer"&gt;Topology and results&lt;/a&gt;.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;OpenVPN (v0.9):&lt;/strong&gt; temporary mutually verified TLS/TUN peers on one kernel test visible/blocked ICMP and a deliberately lower return egress. Small ACKs return at all tested forward sizes, while larger same-sized replies are missing or locally rejected after learned PMTU state. Receiver joins keep those cases separate, and the sized search stays inconclusive. &lt;a href="https://github.com/lazypanic999/tunnelscope/blob/main/docs/validation/openvpn-routed.md" rel="noopener noreferrer"&gt;Feedback validation&lt;/a&gt; and &lt;a href="https://github.com/lazypanic999/tunnelscope/blob/main/docs/validation/openvpn-asymmetric-return.md" rel="noopener noreferrer"&gt;return-path evidence&lt;/a&gt;.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Native Windows receiver / WSL2 sender (v0.9.1):&lt;/strong&gt; over existing IPv4 and IPv6 link-local virtual paths, each family returned all 3 small ACKs, 6 same-sized replies and 8 diagnosis replies with matching receiver observations. A separate 1600-byte comparison then recorded actual Windows &lt;code&gt;WSAEMSGSIZE&lt;/code&gt; (10040): the probes arrived and small ACKs returned, while same-sized reply sends failed locally. These are two kernels on &lt;strong&gt;one physical computer&lt;/strong&gt;, without new firewall rules or MTU changes; physical links and routed IPv6 need separate evidence. &lt;a href="https://github.com/lazypanic999/tunnelscope/blob/main/docs/validation/windows-wsl2-receiver.md" rel="noopener noreferrer"&gt;IPv4&lt;/a&gt;, &lt;a href="https://github.com/lazypanic999/tunnelscope/blob/main/docs/validation/windows-wsl2-receiver-ipv6.md" rel="noopener noreferrer"&gt;IPv6 scopes&lt;/a&gt;, and &lt;a href="https://github.com/lazypanic999/tunnelscope/blob/main/docs/validation/windows-oversized-replies.md" rel="noopener noreferrer"&gt;oversized reply evidence&lt;/a&gt;.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Kubernetes/Calico:&lt;/strong&gt; pods on two Docker-based Minikube nodes each delivered 6/6 baseline probes across sizes 64, 1200 and 1400 bytes. A scoped ingress policy produced small-probe timeouts; restoring the rule restored delivery. Their configured 1480-byte pod interfaces produced a local 1452/1453-byte payload interval. &lt;a href="https://github.com/lazypanic999/tunnelscope/blob/main/docs/validation/kubernetes-calico.md" rel="noopener noreferrer"&gt;Calico IP-in-IP validation&lt;/a&gt;.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;These checks ran on one physical host. Production VPN incidents, physical multi-host paths and other CNI modes need separate evidence. An observed payload interval is not an independently measured end-to-end PMTU. ICMP and receiver logs remain unauthenticated, so TunnelScope is intended for controlled environments.&lt;/p&gt;

&lt;p&gt;The &lt;a href="https://github.com/lazypanic999/tunnelscope" rel="noopener noreferrer"&gt;code and release notes&lt;/a&gt;, &lt;a href="https://github.com/lazypanic999/tunnelscope/blob/main/docs/diagnose.md" rel="noopener noreferrer"&gt;diagnose guide&lt;/a&gt; and &lt;a href="https://github.com/lazypanic999/tunnelscope/blob/main/docs/kubernetes.md" rel="noopener noreferrer"&gt;Kubernetes lab&lt;/a&gt; are available. If you can test a permitted VPN path or reach a Windows receiver from a separate Linux host, the &lt;a href="https://github.com/lazypanic999/tunnelscope/blob/main/docs/field-reports.md" rel="noopener noreferrer"&gt;field report guide&lt;/a&gt; explains bounded collection and sanitization. A normal working path is useful too; no remote access, credentials or private captures need to be shared. &lt;a href="https://github.com/lazypanic999/tunnelscope/issues/2" rel="noopener noreferrer"&gt;Independent operator VPN reports remain welcome&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;Thanks to &lt;a href="https://github.com/sunxiuguo" rel="noopener noreferrer"&gt;sunxiuguo&lt;/a&gt; for the &lt;a href="https://github.com/lazypanic999/tunnelscope/pull/53" rel="noopener noreferrer"&gt;report-kind display fix&lt;/a&gt;, &lt;a href="https://github.com/lazypanic999/tunnelscope/pull/56" rel="noopener noreferrer"&gt;Markdown sharing guide&lt;/a&gt; and &lt;a href="https://github.com/lazypanic999/tunnelscope/pull/66" rel="noopener noreferrer"&gt;sender-kind validation&lt;/a&gt;; and to &lt;a href="https://github.com/HarshRajSinghania" rel="noopener noreferrer"&gt;HarshRajSinghania&lt;/a&gt; for &lt;a href="https://github.com/lazypanic999/tunnelscope/pull/61" rel="noopener noreferrer"&gt;Python 3.13 CI&lt;/a&gt; and &lt;a href="https://github.com/lazypanic999/tunnelscope/pull/67" rel="noopener noreferrer"&gt;receiver scope metadata&lt;/a&gt;. The final changes were reviewed and passed the delivery checks.&lt;/p&gt;

</description>
      <category>showdev</category>
      <category>python</category>
      <category>networking</category>
      <category>linux</category>
    </item>
    <item>
      <title>Building CorePulse: a guarded 5G core fault lab in Go</title>
      <dc:creator>Doğukan Sağlık</dc:creator>
      <pubDate>Sun, 01 Nov 2026 18:15:22 +0000</pubDate>
      <link>https://dev.to/lazypanic999/building-corepulse-a-guarded-5g-core-fault-lab-in-go-4bpf</link>
      <guid>https://dev.to/lazypanic999/building-corepulse-a-guarded-5g-core-fault-lab-in-go-4bpf</guid>
      <description>&lt;p&gt;I am building &lt;a href="https://github.com/lazypanic999/CorePulse" rel="noopener noreferrer"&gt;CorePulse&lt;/a&gt;, an open-source Go project for experimenting with failures in a local 5G core lab. It runs alongside Open5GS and UERANSIM on Kubernetes. The long-term goal is to collect labeled fault telemetry and evaluate detection models. The current version is deliberately earlier: a working control plane, one guarded fault type, live resource charts, and a service-path check.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fybtcmpi0474f42ae3opv.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fybtcmpi0474f42ae3opv.png" alt="CorePulse dashboard connected to the local lab" width="800" height="967"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  Why I started with a controlled experiment
&lt;/h2&gt;

&lt;p&gt;It is easy to draw a fault-detection dashboard before there is a trustworthy fault label. I wanted the first usable slice to answer a simpler question: can I deliberately break one component, record when it happened, and tell whether Kubernetes and the UE path recovered?&lt;/p&gt;

&lt;p&gt;CorePulse currently supports an AMF or UPF pod experiment in the isolated &lt;code&gt;open5gs-lab&lt;/code&gt; context. The default command prints a plan. Applying the experiment checks the exact context, namespace, workload ownership, and readiness; deletes one deployment-owned pod; waits for deployment recovery; and writes a run record. The local dashboard repeats those checks and requires a typed target confirmation. It binds to loopback and disables the run control when the lab is unavailable.&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;go run ./cmd/corepulse pod-crash &lt;span class="nt"&gt;--target&lt;/span&gt; amf
go run ./cmd/corepulse pod-crash &lt;span class="nt"&gt;--target&lt;/span&gt; amf &lt;span class="nt"&gt;--apply&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;h2&gt;
  
  
  What the dashboard actually measures
&lt;/h2&gt;

&lt;p&gt;While the server runs, it samples Kubernetes pod CPU and memory every 10 seconds. &lt;code&gt;corepulse dataset&lt;/code&gt; joins those samples with recorded experiment intervals to produce a labeled CSV. The dashboard also runs a read-only check from a ready UE pod to the UPF tunnel gateway. I keep that result separate from deployment readiness: a replacement pod becoming ready does not prove the UE data path works.&lt;/p&gt;

&lt;p&gt;The first local lab verification passed UE registration, PDU session establishment, and external ping after correcting the UERANSIM gNB N3 address. That setup fix and its verification are documented in the repository. The gateway probe in CorePulse itself tests only UE-to-UPF forwarding; it does not claim to verify external reachability.&lt;/p&gt;

&lt;h2&gt;
  
  
  What is still missing
&lt;/h2&gt;

&lt;p&gt;There is &lt;strong&gt;no trained detector or prediction in this preview&lt;/strong&gt;. The current chart is resource telemetry. The initial CSV includes lab startup and diagnostic activity, so I would not use it as a clean training set. Next I want repeated, isolated runs, more reversible fault types, and a model evaluation split by run rather than by individual sample window.&lt;/p&gt;

&lt;p&gt;The &lt;a href="https://github.com/lazypanic999/CorePulse" rel="noopener noreferrer"&gt;repository&lt;/a&gt; has the setup notes, tests, dashboard screenshots, and &lt;a href="https://github.com/lazypanic999/CorePulse/releases/tag/v0.1.0" rel="noopener noreferrer"&gt;v0.1.0 preview release&lt;/a&gt;. The current &lt;code&gt;main&lt;/code&gt; branch also contains dashboard and gateway-probe changes made after that release.&lt;/p&gt;

&lt;p&gt;If you have worked on telecom labs or chaos experiments, I would appreciate feedback on one design choice: which signal would you prioritize next for distinguishing AMF trouble from UPF trouble without mistaking Kubernetes pod readiness for end-to-end service health?&lt;/p&gt;

&lt;p&gt;&lt;em&gt;Disclosure: This post was drafted with AI assistance. Technical claims were checked against the repository and lab notes.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>showdev</category>
      <category>kubernetes</category>
      <category>go</category>
      <category>devops</category>
    </item>
    <item>
      <title>Looking for contributors to TunnelScope: Python MTU diagnostics with good first issues</title>
      <dc:creator>Doğukan Sağlık</dc:creator>
      <pubDate>Wed, 28 Oct 2026 22:34:29 +0000</pubDate>
      <link>https://dev.to/lazypanic999/looking-for-contributors-to-tunnelscope-python-mtu-diagnostics-with-good-first-issues-179</link>
      <guid>https://dev.to/lazypanic999/looking-for-contributors-to-tunnelscope-python-mtu-diagnostics-with-good-first-issues-179</guid>
      <description>&lt;p&gt;I'm building &lt;a href="https://github.com/lazypanic999/tunnelscope" rel="noopener noreferrer"&gt;TunnelScope&lt;/a&gt;, an MIT-licensed Python CLI for evidence-based MTU and packet-size experiments. Linux senders and Linux or Windows receivers retain observations in JSON, with offline HTML and bounded Markdown for review.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Update — 8 November 2026:&lt;/strong&gt; &lt;a href="https://pypi.org/project/tunnelscope/0.9.3/" rel="noopener noreferrer"&gt;v0.9.3 alpha is on PyPI&lt;/a&gt;. Several starter tasks have now been completed by contributors. The &lt;a href="https://github.com/lazypanic999/tunnelscope/blob/main/docs/demo-v09.md" rel="noopener noreferrer"&gt;30-second recorded evidence replay&lt;/a&gt; shows the sharing workflow.&lt;/p&gt;

&lt;p&gt;TunnelScope is still an alpha tool: a timeout is inconclusive, and an observed payload interval is not an independently measured end-to-end PMTU or proof of a root cause. Its UDP protocol and receiver logs are unauthenticated, so experiments belong on permitted controlled paths.&lt;/p&gt;

&lt;h2&gt;
  
  
  Recent completed contributions
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;a href="https://github.com/lazypanic999/tunnelscope/pull/61" rel="noopener noreferrer"&gt;Python 3.13 delivery coverage&lt;/a&gt;: all six Linux/Windows matrix jobs pass; issue #33 is completed.&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://github.com/lazypanic999/tunnelscope/pull/66" rel="noopener noreferrer"&gt;Reject a joined report used as raw sender input&lt;/a&gt;: preserves legacy inputs and checks that errors create no output files; issue #62 is completed.&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://github.com/lazypanic999/tunnelscope/pull/67" rel="noopener noreferrer"&gt;Retain the IPv6 receiver's local scope&lt;/a&gt;: adds local metadata without treating it as a cross-host identity; issue #64 is completed.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Check the &lt;a href="https://github.com/lazypanic999/tunnelscope/issues?q=is%3Aissue%20is%3Aopen" rel="noopener noreferrer"&gt;live open task list&lt;/a&gt; for current work. A small reproducible bug report or focused documentation improvement is welcome too. Please comment before starting a listed task, and discuss larger changes first. The original Python 3.10 CI, receiver quick-start and PowerShell documentation tasks from this post were completed earlier.&lt;/p&gt;

&lt;h2&gt;
  
  
  Field reports are welcome too
&lt;/h2&gt;

&lt;p&gt;Controlled records cover WireGuard, OpenVPN, IPv6 feedback, Windows/WSL2 host/guest paths and one Kubernetes/Calico setup. Those checks do not replace an &lt;a href="https://github.com/lazypanic999/tunnelscope/issues/2" rel="noopener noreferrer"&gt;independent operator VPN report&lt;/a&gt; or physical multi-host evidence. A normal working path is useful too. The &lt;a href="https://github.com/lazypanic999/tunnelscope/blob/main/docs/field-reports.md" rel="noopener noreferrer"&gt;field guide&lt;/a&gt; explains bounded collection, joining and sanitizing related reports together. Review copies before sharing; no remote access, credentials or private packet captures need to be shared.&lt;/p&gt;

&lt;p&gt;Thanks to &lt;a href="https://github.com/sunxiuguo" rel="noopener noreferrer"&gt;sunxiuguo&lt;/a&gt; for PRs &lt;a href="https://github.com/lazypanic999/tunnelscope/pull/53" rel="noopener noreferrer"&gt;#53&lt;/a&gt;, &lt;a href="https://github.com/lazypanic999/tunnelscope/pull/56" rel="noopener noreferrer"&gt;#56&lt;/a&gt; and &lt;a href="https://github.com/lazypanic999/tunnelscope/pull/66" rel="noopener noreferrer"&gt;#66&lt;/a&gt;, and &lt;a href="https://github.com/HarshRajSinghania" rel="noopener noreferrer"&gt;HarshRajSinghania&lt;/a&gt; for &lt;a href="https://github.com/lazypanic999/tunnelscope/pull/61" rel="noopener noreferrer"&gt;#61&lt;/a&gt; and &lt;a href="https://github.com/lazypanic999/tunnelscope/pull/67" rel="noopener noreferrer"&gt;#67&lt;/a&gt;. Five PRs from two contributors have merged after maintainer review and delivery checks.&lt;/p&gt;

&lt;p&gt;Start with the &lt;a href="https://github.com/lazypanic999/tunnelscope#readme" rel="noopener noreferrer"&gt;README&lt;/a&gt; and &lt;a href="https://github.com/lazypanic999/tunnelscope/blob/main/CONTRIBUTING.md" rel="noopener noreferrer"&gt;contribution guide&lt;/a&gt;. Reproducible reports, documentation feedback and focused PRs are welcome.&lt;/p&gt;

</description>
      <category>contributorswanted</category>
      <category>python</category>
      <category>networking</category>
      <category>beginners</category>
    </item>
  </channel>
</rss>
