<?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: Solution Market</title>
    <description>The latest articles on DEV Community by Solution Market (@solutionmarket).</description>
    <link>https://dev.to/solutionmarket</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%2F4246086%2F6f55c628-5bc0-47f0-bad2-0c107ed039fc.png</url>
      <title>DEV Community: Solution Market</title>
      <link>https://dev.to/solutionmarket</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/solutionmarket"/>
    <language>en</language>
    <item>
      <title>Our MySQL rollup job blocked the sensors it was supposed to summarize</title>
      <dc:creator>Solution Market</dc:creator>
      <pubDate>Mon, 16 Nov 2026 01:25:27 +0000</pubDate>
      <link>https://dev.to/solutionmarket/our-mysql-rollup-job-blocked-the-sensors-it-was-supposed-to-summarize-2fe3</link>
      <guid>https://dev.to/solutionmarket/our-mysql-rollup-job-blocked-the-sensors-it-was-supposed-to-summarize-2fe3</guid>
      <description>&lt;p&gt;&lt;em&gt;InfraGuard Engineering Devlog, part 6 of 6.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;On November 12, five commits went into InfraGuard's rollup worker. Retention cleanup was not completing, a drain path had replayed data, and the worker blocked sensor inserts.&lt;/p&gt;

&lt;p&gt;The common theme was a mismatch between database work and process lifetime.&lt;/p&gt;

&lt;h2&gt;
  
  
  A minute-long process, a seven-minute workflow
&lt;/h2&gt;

&lt;p&gt;PM2 restarted the worker every minute. Database statements could outlive the client, while subsequent script steps never ran.&lt;/p&gt;

&lt;p&gt;One drain operation aggregated a detached traffic copy in a statement taking about seven minutes. MySQL could commit that statement after the client was killed, but the subsequent rename recording progress did not happen. The next execution added the same copy again.&lt;/p&gt;

&lt;p&gt;The engineering record described 53 nights of repeated additions and an undetached raw table accumulating about 58.8 million rows over 54 days.&lt;/p&gt;

&lt;p&gt;We replaced the all-at-once drain with hourly transactions: aggregate one hour, delete that hour from the detached source, and commit. Output and progress now share a transaction. If it rolls back, both roll back. Remaining source data represents remaining work.&lt;/p&gt;

&lt;h2&gt;
  
  
  Budget each run
&lt;/h2&gt;

&lt;p&gt;The worker has a 40-second execution budget and stops starting new work when it is exhausted. Retention cleanup runs in short slices rather than relying on a daily slot that might never finish.&lt;/p&gt;

&lt;p&gt;A budget cannot make an arbitrary in-flight statement instantaneous. Work units must also be small, restartable and durably recorded.&lt;/p&gt;

&lt;h2&gt;
  
  
  The reader blocked writers
&lt;/h2&gt;

&lt;p&gt;Under the deployed MySQL configuration's REPEATABLE READ behavior, &lt;code&gt;INSERT … SELECT&lt;/code&gt; rollups acquired shared locks while scanning source rows. Sensors inserting late data into the recomputed interval waited behind them.&lt;/p&gt;

&lt;p&gt;Recorded ordinary waits were 27–58 seconds every five minutes. Overlapping maintenance pushed a wait to roughly five minutes. Batches from six sensors exceeded the 240-second lock timeout. Connection pools filled the account's 50-connection limit, and failed batches were discarded rather than replayed: silent data loss.&lt;/p&gt;

&lt;p&gt;The worker now changes its own session to READ COMMITTED when the binary-log configuration supports it. The documented configuration requires ROW; STATEMENT retains the existing isolation level and emits a warning. This is not advice to change global isolation blindly.&lt;/p&gt;

&lt;p&gt;A scratch-table reproduction showed a late insert timing out under REPEATABLE READ and passing under READ COMMITTED. Separate per-source, heavy-rollup and maintenance locks also let unrelated sources keep progressing.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Takeaway:&lt;/strong&gt; Make output and progress atomic, bound work to process lifetime, and reproduce locking under the actual isolation and logging configuration. A successful summary is not success if producing it prevents new observations from arriving.&lt;/p&gt;




&lt;p&gt;&lt;strong&gt;About this series:&lt;/strong&gt; Adapted from InfraGuard engineering records. Measurements describe the recorded environment, not product-wide benchmarks. AI was used to draft and translate this article.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Disclosure:&lt;/strong&gt; Solution Market develops and sells InfraGuard. This is our engineering devlog, not an independent review. InfraGuard is a self-hosted infrastructure operations and security platform. &lt;a href="https://infraguard.work" rel="noopener noreferrer"&gt;Project&lt;/a&gt; · &lt;a href="https://infraguard.work/manual" rel="noopener noreferrer"&gt;Documentation&lt;/a&gt; · &lt;a href="https://solutionmarket.net/en" rel="noopener noreferrer"&gt;Solution Market&lt;/a&gt;&lt;/p&gt;

</description>
      <category>sql</category>
      <category>devops</category>
      <category>debugging</category>
      <category>buildinpublic</category>
    </item>
    <item>
      <title>A repaired sensor made our dashboard take 28 seconds to load</title>
      <dc:creator>Solution Market</dc:creator>
      <pubDate>Mon, 16 Nov 2026 01:25:11 +0000</pubDate>
      <link>https://dev.to/solutionmarket/a-repaired-sensor-made-our-dashboard-take-28-seconds-to-load-1548</link>
      <guid>https://dev.to/solutionmarket/a-repaired-sensor-made-our-dashboard-take-28-seconds-to-load-1548</guid>
      <description>&lt;p&gt;&lt;em&gt;InfraGuard Engineering Devlog, part 5 of 6.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;Restoring a stopped sensor on October 30 made telemetry healthier and the console slower. Recorded cold-cache tests averaged 13.3 seconds for the first view and reached 28.5 seconds.&lt;/p&gt;

&lt;p&gt;A weak-TLS aggregation was responsible. Recent 24-hour TLS data had grown to 3.01 million of 11.67 million rows. The optimizer changed its access plan; one query took 13–36 seconds.&lt;/p&gt;

&lt;h2&gt;
  
  
  One covering index helped
&lt;/h2&gt;

&lt;p&gt;An index covering time and TLS version let this query read the index alone. The recorded online build took 82 seconds and added 353 MB. The aggregation then completed in about 1.8 seconds.&lt;/p&gt;

&lt;p&gt;We also precomputed the score every four minutes rather than making the first viewer pay the cold-query cost. Improving a query and removing expensive computation from a request path are separate fixes.&lt;/p&gt;

&lt;h2&gt;
  
  
  Another index did not
&lt;/h2&gt;

&lt;p&gt;An earlier seven-day JA3 top-IP query took 3,728 ms. A covering-index experiment did not help, so we removed it. An index hint was slower at roughly 4–6 seconds.&lt;/p&gt;

&lt;p&gt;An hourly rollup helped that workload: about 20 ms, a recorded improvement of roughly 186 times. “Use rollups for wide windows” sounded like a rule. On November 4, JA3 provided the counterexample.&lt;/p&gt;

&lt;h2&gt;
  
  
  Rollups need something to collapse
&lt;/h2&gt;

&lt;p&gt;After recovery, a 24-hour TLS window held 3.06 million rows and 1.12 million distinct source IPs. Bot traffic kept grouping cardinality high. The rollup retained 2.28 million rows. Its recorded footprint was 7.6 GB, versus 6.3 GB for the raw table.&lt;/p&gt;

&lt;p&gt;Reading the 24-hour rollup took 45–131 seconds. Grouping a million addresses spilled intermediate work to disk. The access path was not the whole problem.&lt;/p&gt;

&lt;p&gt;We capped JA3 aggregation at six recent hours, read raw data, and shared one top-300 snapshot across screens. If the UI selects a broader range, it explicitly says the JA3 result still represents six hours.&lt;/p&gt;

&lt;h2&gt;
  
  
  Cache duration is not a computation budget
&lt;/h2&gt;

&lt;p&gt;A 30-second cache with 30-second polling can make requests arrive just as entries expire. Recomputations taking 23–85 seconds then overlap across tabs on the same database.&lt;/p&gt;

&lt;p&gt;We moved computation behind precomputed snapshots.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Takeaway:&lt;/strong&gt; Inspect the plan, measure the rollup-to-raw row ratio, and test the request path. Indexes, rollups and caches solve different problems. None replaces checking the actual workload.&lt;/p&gt;




&lt;p&gt;&lt;strong&gt;About this series:&lt;/strong&gt; Adapted from InfraGuard engineering records. Measurements describe the recorded environment, not product-wide benchmarks. AI was used to draft and translate this article.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Disclosure:&lt;/strong&gt; Solution Market develops and sells InfraGuard. This is our engineering devlog, not an independent review. InfraGuard is a self-hosted infrastructure operations and security platform. &lt;a href="https://infraguard.work" rel="noopener noreferrer"&gt;Project&lt;/a&gt; · &lt;a href="https://infraguard.work/manual" rel="noopener noreferrer"&gt;Documentation&lt;/a&gt; · &lt;a href="https://solutionmarket.net/en" rel="noopener noreferrer"&gt;Solution Market&lt;/a&gt;&lt;/p&gt;

</description>
      <category>sql</category>
      <category>performance</category>
      <category>database</category>
      <category>buildinpublic</category>
    </item>
    <item>
      <title>We blocked our own DNS resolver: when IDS alerts are not attacker evidence</title>
      <dc:creator>Solution Market</dc:creator>
      <pubDate>Mon, 16 Nov 2026 01:19:43 +0000</pubDate>
      <link>https://dev.to/solutionmarket/we-blocked-our-own-dns-resolver-when-ids-alerts-are-not-attacker-evidence-4dd1</link>
      <guid>https://dev.to/solutionmarket/we-blocked-our-own-dns-resolver-when-ids-alerts-are-not-attacker-evidence-4dd1</guid>
      <description>&lt;p&gt;&lt;em&gt;InfraGuard Engineering Devlog, part 4 of 6.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;On October 25, we found Cloudflare's public resolver, &lt;code&gt;1.1.1.1&lt;/code&gt;, in our own blocking pipeline.&lt;/p&gt;

&lt;p&gt;Repeated IDS activity nominated sources; external reputation helped decide whether to block them. Individual IP blocks could later be promoted to a /24 range when enough distinct addresses accumulated. Staging already blocked individual IPs; promotion expanded the scope.&lt;/p&gt;

&lt;p&gt;The design failed when the counted evidence described our sensor rather than the remote host.&lt;/p&gt;

&lt;h2&gt;
  
  
  Unsolicited from whose perspective?
&lt;/h2&gt;

&lt;p&gt;The signature was “SURICATA DNS Unsolicited response.” The record showed 1,572 events over seven days from just two sources, including &lt;code&gt;1.1.1.1&lt;/code&gt;.&lt;/p&gt;

&lt;p&gt;A separate 30-day inspection found 1,328 records directed to our host's ephemeral UDP ports. These were responses to our queries, but the sensor no longer retained the request state.&lt;/p&gt;

&lt;p&gt;Persistence nominated the resolver. Reputation did not rescue the decision: reports against a shared resolver made a misleading local signal appear corroborated.&lt;/p&gt;

&lt;h2&gt;
  
  
  Visibility failure is not attribution
&lt;/h2&gt;

&lt;p&gt;Some alerts describe packets a remote endpoint sent. Others describe flow tracking, capture-thread distribution or lost context in the sensor.&lt;/p&gt;

&lt;p&gt;Treating the latter as hostile-peer evidence can build the strongest-looking case against normal high-volume infrastructure. More legitimate traffic creates more opportunities for local visibility gaps.&lt;/p&gt;

&lt;p&gt;Excluding the non-attributable signature category reduced candidates from 199 to 183. In that inspection, the removed 16 were CDN or cloud edges, including CloudFront, Amazon and Cloudflare infrastructure.&lt;/p&gt;

&lt;h2&gt;
  
  
  Fix the evidence class
&lt;/h2&gt;

&lt;p&gt;We added a public-resolver allowlist, but an address list cannot fix a category error. The same failure could affect an unfamiliar resolver or CDN edge.&lt;/p&gt;

&lt;p&gt;The durable change was a canonical code-level list of non-attributable signatures, shared by both candidate-selection queries. Whether a signature supports attribution is a property of the evidence, not merely an operator's tuning preference.&lt;/p&gt;

&lt;p&gt;A November 12 change made allowlisting clean up matching existing block records too. “Allowed now” should not leave an earlier block silently effective.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Takeaway:&lt;/strong&gt; Persistence strengthens evidence only after you establish what the evidence is about. Two systems agreeing on a shared false premise is not independent confirmation.&lt;/p&gt;




&lt;p&gt;&lt;strong&gt;About this series:&lt;/strong&gt; Adapted from InfraGuard engineering records. Measurements describe the recorded environment, not product-wide benchmarks. AI was used to draft and translate this article.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Disclosure:&lt;/strong&gt; Solution Market develops and sells InfraGuard. This is our engineering devlog, not an independent review. InfraGuard is a self-hosted infrastructure operations and security platform. &lt;a href="https://infraguard.work" rel="noopener noreferrer"&gt;Project&lt;/a&gt; · &lt;a href="https://infraguard.work/manual" rel="noopener noreferrer"&gt;Documentation&lt;/a&gt; · &lt;a href="https://solutionmarket.net/en" rel="noopener noreferrer"&gt;Solution Market&lt;/a&gt;&lt;/p&gt;

</description>
      <category>security</category>
      <category>networking</category>
      <category>devops</category>
      <category>buildinpublic</category>
    </item>
    <item>
      <title>Our security score punished successful blocking. We changed what it counted.</title>
      <dc:creator>Solution Market</dc:creator>
      <pubDate>Mon, 16 Nov 2026 01:18:46 +0000</pubDate>
      <link>https://dev.to/solutionmarket/our-security-score-punished-successful-blocking-we-changed-what-it-counted-bki</link>
      <guid>https://dev.to/solutionmarket/our-security-score-punished-successful-blocking-we-changed-what-it-counted-bki</guid>
      <description>&lt;p&gt;&lt;em&gt;InfraGuard Engineering Devlog, part 3 of 6.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;A security score should improve when defenses work. On September 4, ours did the opposite.&lt;/p&gt;

&lt;p&gt;The formula deducted points for observed high-severity alerts, anomaly-engine blocks and newly blocked sources. The recorded dataset contained 889 anomaly-engine blocks and 789 newly blocked reputation-based sources that day. Those successes pushed components toward maximum penalties.&lt;/p&gt;

&lt;p&gt;Of 282 high-severity IDS alerts, 227, about 80.5%, came from already blocked ranges. They still counted against the score.&lt;/p&gt;

&lt;h2&gt;
  
  
  Count the unresolved remainder
&lt;/h2&gt;

&lt;p&gt;We replaced volume penalties with measures of what remained unaddressed:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;IDS:&lt;/strong&gt; high-severity alerts from sources not yet blocked, relative to observed high-severity alerts.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Anomaly detection:&lt;/strong&gt; above-threshold findings not blocked by the engine or current blocklist.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Reputation:&lt;/strong&gt; the share of observed attacking source ranges still unblocked, counted by source rather than event volume.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;“The sensor observed something bad” and “we still have actionable exposure” are not interchangeable.&lt;/p&gt;

&lt;p&gt;The formula also uses smoothing. A tiny denominator should not make one unresolved finding send a quiet environment straight to its worst score. At displayed precision, recorded examples were 100 for one unresolved item, 98 for five, and 55 for 50,000 wholly unaddressed items. These are implementation examples, not thresholds to copy without examining the formula and context.&lt;/p&gt;

&lt;h2&gt;
  
  
  One dataset, two answers
&lt;/h2&gt;

&lt;p&gt;Two overview screens implemented different formulas: one displayed 62 and another 55 for the same data. Both now call a shared scoring service.&lt;/p&gt;

&lt;p&gt;We also found an “unaddressed threats” metric that displayed successful blocks. The label described open work; the value counted completed work.&lt;/p&gt;

&lt;p&gt;Recalculating the recorded dataset changed the overall score from 62 to 80. Even complete blocking would have left that snapshot at 94, because separate penalties remained for exposed critical ports and weak TLS settings.&lt;/p&gt;

&lt;h2&gt;
  
  
  Expose the evidence
&lt;/h2&gt;

&lt;p&gt;Each component shows its unresolved and observed counts. If data cannot be read, the application should not invent a score from traffic volume.&lt;/p&gt;

&lt;p&gt;A blocked source also does not prove the absence of compromise. This is an operational prioritization signal, not a security certification.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Takeaway:&lt;/strong&gt; Does a metric's label match what it counts? When the number improves, does the operational outcome improve too? Ask both before putting a score at the center of a dashboard.&lt;/p&gt;




&lt;p&gt;&lt;strong&gt;About this series:&lt;/strong&gt; Adapted from InfraGuard engineering records. Measurements describe the recorded environment, not product-wide benchmarks. AI was used to draft and translate this article.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Disclosure:&lt;/strong&gt; Solution Market develops and sells InfraGuard. This is our engineering devlog, not an independent review. InfraGuard is a self-hosted infrastructure operations and security platform. &lt;a href="https://infraguard.work" rel="noopener noreferrer"&gt;Project&lt;/a&gt; · &lt;a href="https://infraguard.work/manual" rel="noopener noreferrer"&gt;Documentation&lt;/a&gt; · &lt;a href="https://solutionmarket.net/en" rel="noopener noreferrer"&gt;Solution Market&lt;/a&gt;&lt;/p&gt;

</description>
      <category>security</category>
      <category>devops</category>
      <category>analytics</category>
      <category>buildinpublic</category>
    </item>
    <item>
      <title>Why a weekly backup kept rebuilding a 5 GB archive</title>
      <dc:creator>Solution Market</dc:creator>
      <pubDate>Mon, 16 Nov 2026 01:12:16 +0000</pubDate>
      <link>https://dev.to/solutionmarket/why-a-weekly-backup-kept-rebuilding-a-5-gb-archive-dj6</link>
      <guid>https://dev.to/solutionmarket/why-a-weekly-backup-kept-rebuilding-a-5-gb-archive-dj6</guid>
      <description>&lt;p&gt;&lt;em&gt;InfraGuard Engineering Devlog, part 2 of 6.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;On August 18, a weekly backup was repeatedly creating a roughly 5 GB full archive. The cause was a chain of process, reporting and scheduling failures.&lt;/p&gt;

&lt;h2&gt;
  
  
  Five links in the failure chain
&lt;/h2&gt;

&lt;ol&gt;
&lt;li&gt;PM2 restarted the backup runner every minute through &lt;code&gt;cron_restart&lt;/code&gt;.&lt;/li&gt;
&lt;li&gt;When the runner died, its child backup shell lost its output pipe.&lt;/li&gt;
&lt;li&gt;A subsequent log write could terminate the shell with &lt;code&gt;SIGPIPE&lt;/code&gt;, before it reported a result.&lt;/li&gt;
&lt;li&gt;With no report, the application did not advance the next-run timestamp.&lt;/li&gt;
&lt;li&gt;The next runner tick still saw an overdue job and started it again.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;Redirecting errors or adding &lt;code&gt;|| true&lt;/code&gt; did not solve signal-driven termination. Swallowing an exit status and surviving a signal are different things.&lt;/p&gt;

&lt;h2&gt;
  
  
  Change the scheduling trade-off explicitly
&lt;/h2&gt;

&lt;p&gt;We changed eligibility to a phase-based minute slot derived from the start time and interval, rather than an indefinitely overdue next-run value. A runner missing its eligible minute skips that occurrence instead of repeatedly catching up.&lt;/p&gt;

&lt;p&gt;That trades automatic recovery of a missed slot for avoiding this runaway replay path. Slot eligibility is not a general exactly-once guarantee across multiple runners; concurrency control remains separate.&lt;/p&gt;

&lt;p&gt;The scripts ignore &lt;code&gt;SIGPIPE&lt;/code&gt; and attempt failure reports for handled termination signals such as TERM, INT and HUP.&lt;/p&gt;

&lt;h2&gt;
  
  
  Successful work, rejected report
&lt;/h2&gt;

&lt;p&gt;A second incident produced a 2.1 GB incremental archive but zero execution records. A preview script had been rendered before the job existed, then saved with &lt;code&gt;JOB_ID=0&lt;/code&gt;. Its report was rejected with HTTP 422.&lt;/p&gt;

&lt;p&gt;Without &lt;code&gt;--fail&lt;/code&gt;, curl can exit successfully after an HTTP error response. We had confused transport completion with application acceptance.&lt;/p&gt;

&lt;p&gt;The fix was to render with the real ID before saving, use &lt;code&gt;curl --fail&lt;/code&gt;, and log rejected reports. Retrying an unchanged invalid payload would not make it valid.&lt;/p&gt;

&lt;h2&gt;
  
  
  Tool-specific semantics matter
&lt;/h2&gt;

&lt;p&gt;GNU tar's exit code 1 can mean files changed during archiving. We treated that documented case as a warning, while retaining failure handling for a nonzero database-dump result. A single generic exit-code policy was wrong.&lt;/p&gt;

&lt;p&gt;The manifest workflow also moved preparation ahead of the expensive transfer and used a rename to finalize the validated result. Previously, completed archives had failed to establish their expected incremental baseline.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Takeaway:&lt;/strong&gt; A backup is a schedule, a process tree, a baseline and an accepted result record, not just an archive. Test the boundaries between them.&lt;/p&gt;




&lt;p&gt;&lt;strong&gt;About this series:&lt;/strong&gt; Adapted from InfraGuard engineering records. Measurements describe the recorded environment, not product-wide benchmarks. AI was used to draft and translate this article.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Disclosure:&lt;/strong&gt; Solution Market develops and sells InfraGuard. This is our engineering devlog, not an independent review. InfraGuard is a self-hosted infrastructure operations and security platform. &lt;a href="https://infraguard.work" rel="noopener noreferrer"&gt;Project&lt;/a&gt; · &lt;a href="https://infraguard.work/manual" rel="noopener noreferrer"&gt;Documentation&lt;/a&gt; · &lt;a href="https://solutionmarket.net/en" rel="noopener noreferrer"&gt;Solution Market&lt;/a&gt;&lt;/p&gt;

</description>
      <category>devops</category>
      <category>linux</category>
      <category>debugging</category>
      <category>buildinpublic</category>
    </item>
    <item>
      <title>Rebuilding a legacy Laravel monitoring app without breaking its collectors</title>
      <dc:creator>Solution Market</dc:creator>
      <pubDate>Mon, 16 Nov 2026 01:10:15 +0000</pubDate>
      <link>https://dev.to/solutionmarket/rebuilding-a-legacy-laravel-monitoring-app-without-breaking-its-collectors-1bbb</link>
      <guid>https://dev.to/solutionmarket/rebuilding-a-legacy-laravel-monitoring-app-without-breaking-its-collectors-1bbb</guid>
      <description>&lt;p&gt;&lt;em&gt;InfraGuard Engineering Devlog, part 1 of 6.&lt;/em&gt;&lt;/p&gt;

&lt;p&gt;Rewriting an application is one problem. Rewriting it while a fleet keeps writing into its database is another.&lt;/p&gt;

&lt;p&gt;InfraGuard grew out of an internal Laravel 8 monitoring project. Node.js collectors were submitting server and network telemetry; backup scripts were already running against the existing schema. We could replace the application, but we could not casually rename the tables beneath those processes.&lt;/p&gt;

&lt;h2&gt;
  
  
  New application, old data contract
&lt;/h2&gt;

&lt;p&gt;The legacy review recorded hard-coded credentials in 77 places across 36 files. It also identified raw SQL assembled from user input, plaintext remember-me cookies, and embedded cryptographic or OAuth secrets. The replacement started on Laravel 13.8 and PHP 8.4, with one constraint: leave existing table names, columns and structure intact.&lt;/p&gt;

&lt;p&gt;The database was an integration contract, not merely private storage. A cleaner name in a migration could break a collector that was working perfectly well.&lt;/p&gt;

&lt;p&gt;New application-owned tables use a literal &lt;code&gt;infr_&lt;/code&gt; prefix. We did not enable a connection-wide Laravel table prefix, because that connection still needs to read legacy tables without it. The documented repository snapshot contains 88 migrations following this coexistence rule.&lt;/p&gt;

&lt;h2&gt;
  
  
  Put the reason next to the rule
&lt;/h2&gt;

&lt;p&gt;“Never hard-code credentials” is easy to agree with and easy to regress. Our notes pair rules with reasons and repeatable checks. The useful outcome is an executable validation, not something someone must remember forever.&lt;/p&gt;

&lt;p&gt;This matters when AI coding tools work on the repository. Plausible code can still violate an integration boundary. Explicit constraints and regression checks make that boundary inspectable.&lt;/p&gt;

&lt;p&gt;We also standardized terminology: “domain” means a web hostname; responsibility boundaries use “role” or “area.” Ambiguous labels can become ambiguous permissions and navigation.&lt;/p&gt;

&lt;h2&gt;
  
  
  Preserve the operational story
&lt;/h2&gt;

&lt;p&gt;The goal was a shorter path from diagnosis to response, with prepared installation material and a recovery path when something fails. Monitoring, detection, blocking, backup and restore verification belong to the same operational story.&lt;/p&gt;

&lt;p&gt;This series follows six incidents from that rebuild. Next: a weekly backup repeatedly recreating a roughly 5 GB full archive.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Takeaway:&lt;/strong&gt; Identify which “internal” structures are contracts with other running systems. Modernizing around those contracts can be safer than modernizing through them.&lt;/p&gt;




&lt;p&gt;&lt;strong&gt;About this series:&lt;/strong&gt; Adapted from InfraGuard engineering records. Measurements describe the recorded environment, not product-wide benchmarks. AI was used to draft and translate this article.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Disclosure:&lt;/strong&gt; Solution Market develops and sells InfraGuard. This is our engineering devlog, not an independent review. InfraGuard is a self-hosted infrastructure operations and security platform. &lt;a href="https://infraguard.work" rel="noopener noreferrer"&gt;Project&lt;/a&gt; · &lt;a href="https://infraguard.work/manual" rel="noopener noreferrer"&gt;Documentation&lt;/a&gt; · &lt;a href="https://solutionmarket.net/en" rel="noopener noreferrer"&gt;Solution Market&lt;/a&gt;&lt;/p&gt;

</description>
      <category>laravel</category>
      <category>devops</category>
      <category>buildinpublic</category>
      <category>architecture</category>
    </item>
    <item>
      <title>Meet Solution Market: six full-source products you can run on your own servers</title>
      <dc:creator>Solution Market</dc:creator>
      <pubDate>Sun, 15 Nov 2026 14:40:51 +0000</pubDate>
      <link>https://dev.to/solutionmarket/meet-solution-market-six-full-source-products-you-can-run-on-your-own-servers-12nf</link>
      <guid>https://dev.to/solutionmarket/meet-solution-market-six-full-source-products-you-can-run-on-your-own-servers-12nf</guid>
      <description>&lt;p&gt;Disclosure: this guide is published by Solution Market about our own commercial products. It is a product overview, not an independent review.&lt;/p&gt;

&lt;p&gt;The question behind these six products is practical: when should a team rent a service, build from scratch, or run an existing application on its own infrastructure? The answer depends on the workflow, the people available to operate it, and how much control the business needs over its data and code.&lt;/p&gt;

&lt;p&gt;Solution Market combines a directory of 1,634 free self-hosted solutions with six commercial, full-source products that we build and run. The directory and the commercial catalog are separate: discovering a free project does not mean our commercial products are open source. Here is a technical, use-case-led tour of the six products.&lt;/p&gt;

&lt;h2&gt;
  
  
  1. MN1 Ad Platform: connect advertisers, publishers and operators
&lt;/h2&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%2Fc4wv9z6thjnq21rd9nmx.webp" 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%2Fc4wv9z6thjnq21rd9nmx.webp" alt="Advertisers launch their first campaign within 15 minutes" width="800" height="500"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;An ad business is more than a banner slot. Advertisers need campaigns and reporting; publishers need inventory and integration; operators need review, wallets and settlement. MN1 brings those three sides into one self-hosted product for communities, media sites and niche advertising networks.&lt;/p&gt;

&lt;p&gt;Its serving and collection planes are separated. A publisher-side script serves ads and records impressions and clicks, while background jobs aggregate the data. The listed stack includes Laravel, Vue, MySQL, Redis and a Node collector. AI-assisted copy and banner creation and semantic matching are part of the feature set.&lt;/p&gt;

&lt;p&gt;The implementation detail to check before deployment is money flow: the delivered top-up flow uses bank transfers with operator confirmation. Card-payment integration is a customization, not something to assume from a campaign dashboard. Test campaign approval, impression collection and monthly settlement as one end-to-end workflow.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Full-source list price: $2,350 one-time.&lt;/strong&gt; &lt;a href="https://solutionmarket.net/en/solutions/mn1" rel="noopener noreferrer"&gt;Product and documentation&lt;/a&gt; · &lt;a href="https://demo.mn1.net/" rel="noopener noreferrer"&gt;Try the demo&lt;/a&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  2. SpaceOS Rental Billing: turn contracts and meter readings into invoices
&lt;/h2&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%2Fjwntt98azu5iu3uvxqwf.webp" 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%2Fjwntt98azu5iu3uvxqwf.webp" alt="Spaces, contracts and billing at a glance" width="800" height="500"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Rental operations often start in a spreadsheet, but a space, a tenant, a contract and a monthly invoice are different records with different lifecycles. This product brings them together for building managers, coworking spaces, short stays and hourly rentals. Its dashboards adapt to the chosen business type.&lt;/p&gt;

&lt;p&gt;Operators can map spaces in a floor-plan editor, manage contracts, collect meter readings and generate the monthly invoice batch with an admin action. PDF generation and email delivery are part of that workflow. Bookings, guests and housekeeping extend it beyond long-term leases.&lt;/p&gt;

&lt;p&gt;The stack includes Laravel, Vue, MySQL, Redis, Reverb for real-time updates and DomPDF. Payment, messaging, e-signature and tax-invoice adapters provide integration points, but availability and regional requirements still need checking. A useful evaluation is to take one real billing period, including a utility allocation and a payment correction, through the demo.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Full-source list price: $1,350 one-time.&lt;/strong&gt; &lt;a href="https://solutionmarket.net/en/solutions/landfee" rel="noopener noreferrer"&gt;Product and documentation&lt;/a&gt; · &lt;a href="https://demo.spaceos.work/" rel="noopener noreferrer"&gt;Try the demo&lt;/a&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  3. AutoTraders: make the signal-to-order decision inspectable
&lt;/h2&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%2Fyeia03h144cz66qoedyl.webp" 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%2Fyeia03h144cz66qoedyl.webp" alt="Positions, P&amp;amp;L, risk and the AI market outlook on one dashboard" width="800" height="500"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Automation is easy to oversell in trading. The useful engineering question is not whether an AI sounds confident, but whether you can inspect why an order was accepted or rejected. AutoTraders implements a trigger → AI review → approved order → result logging flow, with recorded rationales.&lt;/p&gt;

&lt;p&gt;The official specification lists 12 exchange and broker adapters, Pine Script compilation, backtesting and simulation. Laravel and Vue sit above MySQL and Redis queues, with Reverb broadcasts for live updates. Multiple AI-provider integrations allow provider choice rather than a single mandatory model.&lt;/p&gt;

&lt;p&gt;Start with simulation, inspect the decision and fill logs, and review API permissions and risk controls before any live use. Backtest results and AI approval do not establish future performance. This is software for implementing and inspecting strategies, not financial advice or a promise of investment returns.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Full-source list price: $1,550 one-time.&lt;/strong&gt; &lt;a href="https://solutionmarket.net/en/solutions/vip-trading" rel="noopener noreferrer"&gt;Product and documentation&lt;/a&gt; · &lt;a href="https://demo.autotraders.kr/" rel="noopener noreferrer"&gt;Try the demo&lt;/a&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  4. DMINX Crypto Exchange: understand execution before evaluating the interface
&lt;/h2&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%2Fyu5h1tdk28i0w7jj0zlu.webp" 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%2Fyu5h1tdk28i0w7jj0zlu.webp" alt="Order book and fills move in real time" width="800" height="688"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;DMINX combines spot and perpetual trading interfaces, convert, copy trading, Earn, rewards and referrals in one codebase. It also includes operational controls such as KYC submission, TOTP two-factor authentication, withdrawal whitelists, administrator approvals and audit logs.&lt;/p&gt;

&lt;p&gt;An important boundary is explicit in the technical specification: orders are filled on the operator's own books against an external mark price. The displayed bid/ask comes from the feed; this is not a user-to-user matching engine. The operator supplies the external price source. That distinction matters more than how convincing a trading screen looks.&lt;/p&gt;

&lt;p&gt;The stack includes Laravel, Vue, MySQL, Redis queues and Reverb. Evaluate the order, balance, ledger, deposit and withdrawal flows together, and assess price-feed dependencies and reconciliation. Buying source code does not supply an exchange license, legal authorization, liquidity or regulatory compliance. Those responsibilities remain with the operator.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Full-source list price: $4,750 one-time.&lt;/strong&gt; &lt;a href="https://solutionmarket.net/en/solutions/coin-market" rel="noopener noreferrer"&gt;Product and documentation&lt;/a&gt; · &lt;a href="https://demo.dminx.com/" rel="noopener noreferrer"&gt;Try the demo&lt;/a&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  5. menuro Digital Menu: one menu dataset, several service surfaces
&lt;/h2&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%2Fjspcmi1lkjvzv6hnij99.webp" 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%2Fjspcmi1lkjvzv6hnij99.webp" alt="Today's sales, orders and kitchen status on one screen" width="800" height="500"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;A restaurant should not need separate versions of the same menu for QR ordering, a tablet, a kiosk, a digital board and a printed menu. menuro uses shared menu data across these surfaces and sends orders to kitchen and administration screens in real time.&lt;/p&gt;

&lt;p&gt;The product covers menus and options, sold-out items, time-of-day menus, seating, reservations and waitlists. Headquarters and store roles support multi-store operation. AI menu translation supports 24 content languages, while the ordering interface has eight language options. These are different scopes, not a claim that every screen is translated into 24 languages.&lt;/p&gt;

&lt;p&gt;The stack includes Laravel, Vue, MySQL, Redis and Reverb; headless Chrome produces PDF menus. A practical demo test is to change an option or mark an item unavailable, then follow the guest order through the kitchen and payment workflow. Hardware, payment-provider configuration and ongoing operations still need planning.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Full-source list price: $990 one-time.&lt;/strong&gt; &lt;a href="https://solutionmarket.net/en/solutions/menuro" rel="noopener noreferrer"&gt;Product and documentation&lt;/a&gt; · &lt;a href="https://demo.menuro.net/" rel="noopener noreferrer"&gt;Try the demo&lt;/a&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  6. InfraGuard Server Security: connect monitoring, detection, response and recovery
&lt;/h2&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%2F0g6tlgq4ok9uifcjfz9j.webp" 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%2F0g6tlgq4ok9uifcjfz9j.webp" alt="Availability and security scores on one screen" width="800" height="688"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;InfraGuard is for teams that want server telemetry, security events, blocking and backup operations in one self-hosted console. It combines server load, storage and traffic views with Suricata detections, reputation checks, automatic blocking rules and backup-result tracking.&lt;/p&gt;

&lt;p&gt;Its listed architecture includes Laravel, Vue, MySQL, Redis, Node/pm2 agents and a PHP health agent. The detection and response toolchain includes Suricata EVE, conntrack, firewalld, iptables, ipset and AbuseIPDB. The product also includes an in-house Kitsune autoencoder-ensemble implementation for traffic anomaly scoring.&lt;/p&gt;

&lt;p&gt;Treat response automation as an operational system, not a substitute for security judgment. Before enabling automatic blocking, review whitelists, thresholds and external reputation dependencies; verify that collectors are fresh and rehearse a restore. Self-hosting gives control of the deployment and data, while also making upgrades, alert quality and recovery your responsibility.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Full-source list price: $1,780 one-time.&lt;/strong&gt; &lt;a href="https://solutionmarket.net/en/solutions/infraguard" rel="noopener noreferrer"&gt;Product and documentation&lt;/a&gt; · &lt;a href="https://demo.infraguard.work/" rel="noopener noreferrer"&gt;Try the demo&lt;/a&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  What full source means, and what it does not
&lt;/h2&gt;

&lt;p&gt;Each listing offers a one-time delivery of a repository archive with commit history, a base database dump and an installation guide. Secrets and customer operating data are not included. You install and operate the application on your own servers and can modify it for your own business.&lt;/p&gt;

&lt;p&gt;This is a perpetual, non-exclusive, non-transferable commercial license, not a transfer of copyright and not an open-source license. You may provide a service that you operate to your own customers. You may not resell, redistribute or publish the original or modified source, or hand it to another customer to install on their servers. Read the source license before making a deployment or business-model decision.&lt;/p&gt;

&lt;p&gt;The one-time source price is not the total cost of running a service. Budget for hosting, database backups, upgrades, monitoring, and any AI, messaging, exchange or payment-provider usage your deployment needs. Check the support scope, especially if you intend to modify the delivered version.&lt;/p&gt;

&lt;h2&gt;
  
  
  Choose by workflow, then prove the fit in a demo
&lt;/h2&gt;

&lt;p&gt;Start with the business record that matters: a campaign for MN1, a contract and invoice for SpaceOS, an auditable strategy decision for AutoTraders, an order and ledger entry for DMINX, a menu and kitchen order for menuro, or a server event and recovery path for InfraGuard.&lt;/p&gt;

&lt;p&gt;Next, run a complete workflow in the linked demo, read the documentation, and list the integrations and operating responsibilities your team will own. If an existing free solution already fits, the directory is a sensible place to start. If a delivered application fits better, inspect the source-license terms and support boundaries before buying.&lt;/p&gt;

&lt;p&gt;Explore the official catalog and demos at Solution Market. Product features and USD list prices in this guide were checked against the official English listings on 15 November 2026 and may change. Finance and crypto products require a separate assessment of legal, security and market risk.&lt;/p&gt;

&lt;p&gt;Editorial note: prepared with AI assistance using the official product listings and license linked here.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://solutionmarket.net/en" rel="noopener noreferrer"&gt;Explore Solution Market&lt;/a&gt; · &lt;a href="https://solutionmarket.net/en/free" rel="noopener noreferrer"&gt;Free solutions&lt;/a&gt; · &lt;a href="https://solutionmarket.net/en/solutions" rel="noopener noreferrer"&gt;Commercial catalog&lt;/a&gt; · &lt;a href="https://solutionmarket.net/en/legal/source_license_use" rel="noopener noreferrer"&gt;Source license&lt;/a&gt;&lt;/p&gt;

</description>
      <category>selfhosted</category>
      <category>webdev</category>
      <category>abotwrotethis</category>
      <category>laravel</category>
    </item>
  </channel>
</rss>
