Oracle tried to take MySQL from its creator. So he forked it and named it after his other daughter. This is one of the most dramatic stories in open source history. Michael Widenius created MySQL in 1995 and named it after his daughter Mai. In 2008 Sun Microsystems bought it for $1 billion. One year later Oracle bought Sun. The same Oracle whose main product directly competes with MySQL. The same day the acquisition was announced, Michael forked MySQL into MariaDB. Named after his other daughter Maria. He tried to block Oracle entirely. The EU approved the deal anyway. Oracle promised to keep MySQL alive. They technically did. But the best features quietly moved behind a paid enterprise version. Bug reports piled up. Community contributions dried up. The soul of it was gone. MariaDB filled the gap fast. → Fully compatible with MySQL. Zero rewrites needed to migrate. → Wikipedia migrated. → Google migrated. → Most Linux distributions dropped MySQL and made MariaDB the default. Then in 2024 MariaDB was bought by private equity. A different villain. The same feeling. Michael also has a son. He named a database after him too. MaxDB. He only has three kids. So there are no more databases left to name. No more forks left to make. The man who named two databases after his children to protect them from corporations watched both get acquired anyway. Some stories do not have a happy ending. They just have a really good dad.

If you're learning AWS, start with these 30 services. You don't need to master all 200 AWS services. Focus on: - Compute (EC2, Lambda, ECS, EKS) - Storage (S3, EBS, FSx) - Networking (VPC, Route 53, CloudFront) - Databases (RDS, DynamoDB, Aurora) - Monitoring (CloudWatch, X-Ray, CodePipeline) These are the services you'll see most often in real-world projects and interviews. Bookmark this for later. ☁️

4 MCP servers every DevOps engineer should know: 1⃣ Kubernetes MCP: - Investigate CrashLoopBackOff pods - Debug failed deployments - Analyze cluster health Repo: github.com/Flux159/mcp-serve… 2⃣ AWS MCP - Investigate AWS cost spikes - Find unused resources - Troubleshoot cloud infrastructure Repo: github.com/awslabs/mcp 3⃣ Terraform MCP - Review Terraform plans - Detect infrastructure drift - Explain infrastructure changes Repo: github.com/hashicorp/terrafo… 4⃣ Grafana Prometheus MCP - Investigate latency spikes - Analyze production incidents - Explain alert storms Repos: github.com/grafana/mcp-grafa… github.com/pab1it0/prometheu… The interesting part isn't MCP itself. It's that your AI can now access your infrastructure, understand what's happening, and help troubleshoot issues using real data instead of guesses.

Microsoft just made Docker Desktop optional on Windows. WSL Containers was announced at Build 2026 and it is coming to public preview this month. Linux containers will now run natively inside WSL. No Docker Desktop. No third party runtime. No background service eating your RAM. Just WSL. Here is what it actually does: → Run any OCI compatible Linux container directly on Windows through WSL → A new CLI called wslc.exe ships with the next WSL update automatically. Same syntax as Docker so zero learning curve. → A full developer API lets Windows apps run Linux containers silently in the background without the user ever touching a terminal → Enterprise ready out of the box. Works with MBE, Intune and every enterprise management tool your IT team already uses → No separate installation. It arrives as part of your next regular WSL update. WSL started as a way to run Linux tools on Windows. With WSL Containers the line between the two keeps getting thinner. Docker Desktop costs $21 per month per developer for commercial use. WSL Containers is built into Windows and costs nothing. Microsoft is not just making Windows more Linux friendly. They are quietly making every reason to leave Windows for Linux a little harder to justify. Full details here: blogs.windows.com/windowsdev…

Dockerfile best practices every engineer should know. 🐳

GitHub just killed YAML hell. You can now write your GitHub Actions in plain English. This is GitHub Agentic Workflows. Just dropped into public preview. You describe what you want in a plain English Markdown file. GitHub compiles it into Actions YAML automatically. The agent runs it. Here is what it actually does: → Automates reasoning-based tasks like issue triage, CI failure analysis and documentation updates inside GitHub Actions → Agents run with read-only permissions by default inside a sandboxed container → An Agent Workflow Firewall sits between the agent and your codebase → Threat detection scans every proposed change before it gets applied → Outputs go through a safe outputs validation process before anything is merged → Reuses your existing runner groups and policy constraints. Nothing new to configure. Marks & Spencer's CTO said what used to take hours of engineering work now runs autonomously in minutes. The part that matters most for every engineering team is the trust layer. Getting an agent to open a pull request was never the hard part. GitHub built the security architecture that makes it safe enough to actually merge. This is what AI in your CI/CD pipeline looks like when it is done properly. Full announcement here: github.blog/changelog/2026-0…

Vacation is just working from a prettier location.

Nobody wanted PostgreSQL in 1994. Now nobody can live without it. In 1986 Professor Michael Stonebraker at UC Berkeley spent eight years building a database called POSTGRES. Funded by DARPA. Dozens of papers. A team of brilliant students. Then Berkeley shut it down in 1994. Too much maintenance. Not enough research value. The code was left on a server and forgotten. Two graduate students named Andrew Yu and Jolly Chen were not ready to let it die. They took the abandoned code, added SQL support and released it to the world for free. No funding. Just two students who believed the database was worth saving. They called it Postgres95. Two years later it was renamed PostgreSQL. A global community of volunteers took it from there and never stopped. Today PostgreSQL is the most advanced open source database on earth. Notion runs on it. Shopify runs on it. Instagram was built on it. Apple uses it internally. Amazon built Aurora PostgreSQL because their customers demanded it. Companies built on a PostgreSQL base have generated over $2.6 billion in acquisitions. Stonebraker won the Turing Award in 2014. The Nobel Prize of computer science. Partly for the database his own university threw away. Oracle charges enterprises hundreds of thousands of dollars a year for what two grad students rescued from a forgotten server and gave away for free. Some things are too important to let die.

The easiest way to run Linux just dropped and nobody is talking about it. 😳 It is called Webtop. One Docker container. Full Linux desktop. Open it in any browser on any device. No VM. No dual boot. No installation. Just a tab. - KDE, XFCE, MATE and i3 all supported - Ubuntu, Alpine, Arch, Fedora and Debian. Pick your distro. - GPU accelerated via Wayland. Fast enough to actually use daily. - Works on your phone, tablet, Chromebook, anything with a browser. - Full terminal with sudo access inside the container - Share your session live with someone else in real time - Your data persists across restarts like a real machine 4.2k stars. GPL-3.0. Free forever. Chromebook users, Windows users, and developers who want a throwaway dev environment they can open from anywhere, this is for you. Repo here: github.com/linuxserver/docke…

AWS just declared war on Intel and AMD. Graviton5 is now live. This is EC2 M9g and M9gd. The most powerful and energy-efficient chip AWS has ever built. Available right now. Not benchmarks. Real companies on real production traffic: → HubSpot ran MySQL on it. Query duration dropped 60%. Zero code changes. → ClickHouse saw 36% performance boost over Graviton4. Zero code changes. → Honeycomb ran a 6 month production A/B test. 36% better throughput per core. → Meta is deploying tens of millions of Graviton5 cores for agentic AI right now. What makes Graviton5 different: → 25% better compute performance than Graviton4 → 5x larger L3 cache than the previous generation → Fastest DDR5 memory of any processor in the cloud today → 192 cores on the largest instance size → First formally verified cloud hypervisor ever built Intel and AMD have dominated cloud compute for decades. AWS started from scratch 8 years ago. Built 5 generations of custom silicon. And just lapped both of them. Full details here: aws.amazon.com/ec2/instance-…

6 Kubernetes security concepts every engineer should know. 🔒

AWS just replaced your FinOps team with an AI agent. This is AWS FinOps Agent. Just went into public preview. Every cloud team has lived this nightmare. Your AWS bill jumps. Someone gets paged. You spend hours crawling through Cost Explorer, CloudTrail and tag data trying to find the one engineer who spun up something they forgot about. AWS FinOps Agent kills that entire workflow. Here is what it actually does: → Detects a cost anomaly and immediately investigates the root cause without anyone touching it → Correlates the cost spike with CloudTrail events to find exactly what changed and who owns it → Opens a Jira ticket or sends a Slack message directly to the engineer responsible with full context attached → Answers natural language questions like "why did my AWS bill go up last month" using your actual usage data → Generates weekly, monthly and daily cost reports in PDF, HTML or PPT on a schedule automatically → Pulls optimization recommendations from Cost Optimization Hub and Compute Optimizer into one Jira ticket Workday, Convera and AVIV Group are already using it in production. Workday said what used to take hours of manual dashboard work now starts from a natural language interface. Cloud cost management has always been reactive, manual and painful. AWS just made it autonomous. Full details here: aws.amazon.com/finops-agent

A GitLab engineer ran sudo rm -rf on the wrong directory and wiped 300GB of production data. Every single backup failed. In 2016 GitLab was under a heavy spam attack. An engineer tried to fix a replication issue and ran rm -rf on what he thought was the secondary database directory. It was the primary. 300 gigabytes gone. 100,000 users offline. The entire platform down. They went for their backups. → Regular backups were silently failing. Nobody knew. → Volume snapshots were broken. → Azure backups had never been tested. Not once. → S3 backups only covered a different database type. 0 for 4. Every safety net gone. The only thing left was a staging server snapshot. Six hours old. They used it. It worked. But six hours of issues, merge requests and comments were gone forever. Then they did something nobody expected. They live streamed the entire recovery. 5,000 people watched a GitLab engineer try to bring the platform back from the dead in real time. After it was over GitLab published a full postmortem. Every failure. Every mistake. Every lesson. One of the most transparent incident reports in tech history. Most companies would have buried this. GitLab put it on the internet for everyone to read. Remember this the next time you skip testing your backups. If you have never restored from a backup, you do not have backups. You have hopes and dreams.

Certified Kubernetes Autist

Linus Torvalds could have been richer than Elon Musk. He chose not to be. In 2005 his team lost access to the tool they used to manage Linux code overnight. A developer had reverse-engineered it and the company behind it cut them off without warning. Thousands of developers. No way to collaborate. No backup plan. Torvalds did not panic. He sat down and built his own version control system from scratch. In 10 days. He called it Git. On day one it was already tracking its own source code. Within weeks it was managing the entire Linux project. By end of 2005 Git 1.0 was officially released. Then he gave it away for free. Open source. No company. No patents. No monetization. He handed the project off after a few months and went back to working on Linux like nothing happened. Other people saw what he left on the table. GitHub built on top of it. 100 million developers. Microsoft bought it in 2018 for $7.5 billion. GitLab went public in 2021 at nearly $12 billion. Today Git controls over 85% of the version control market. Every app on your phone. Every website you visit. Built using Git. Torvalds made $0 from any of it. He built the most used developer tool in history because he was annoyed. Then gave it away because he believed it should be free for everyone. And he has never once said he regrets it.

Helm cheat sheet for DevOps engineers 🚢

GitHub went from 1 billion commits a year to 1.4 billion commits a month. AI did that. And GitHub is breaking because of it. GitHub COO Kyle Daigle just went public and admitted it openly. The platform that hosts 100 million developers has logged hundreds of incidents in the past 12 months. Outages hitting search, GitHub Actions, CI/CD pipelines. Public apologies. All of it. AI agents alone are generating 17 million pull requests every single month. GitHub planned for 100% growth. The actual number blew past that before they could even react. Here is what they are doing to survive it: → Rebuilding how GitHub Actions dispatches jobs entirely from scratch → Moving performance sensitive code out of their Ruby monolith into Go → Migrating fully off their own data centres onto Azure → Isolating critical services so one broken system cannot take everything else down → Bringing in Microsoft engineers who have scaled systems at this level before The goal is to get GitHub to handle 30 times its current load. Not 30 percent. 30 times. GitHub launched Copilot in 2021 and created the AI coding era. That same wave is now threatening to break their own platform. The company that started the fire is now racing to stop it from burning everything down. Full story here: thenewstack.io/github-wants-…

Same job title. Completely different mental illnesses. 🤪