DevOpsLinks

Most DevOps teams added more tools in 2026 than they shipped features. Nobody wants to say it out loud.

But your "modern stack" is just technical debt with better branding.

Here's what I keep seeing across teams:

• Multiple monitoring tools, zero actionable alerts
• Multiple CI/CD pipelines, none fully owned by anyone
• An internal developer platform that's really just a wiki with links

And the pattern is always the same: new tools are adopted to fix a gap. Nobody decommissions the old one and 6 months later, you're paying for both and trusting neither.

I review hundreds of DevOps tools every year for FAUN•dev. You're subscribed to our newsletters and you started probably to notice which tools keep showing up in real stacks - and which ones disappear after the hype cycle. A few tools always make it through the filter. Not because they're trendy, but because they solve a real problem without creating three new ones.

The pattern is always the same: the fastest teams don't have the most tools. They have the fewest, and they can explain why each one is there.

The 2026 shift isn't about adding AI agents to your workflow. It's about having a workflow clean enough for AI to actually help.

If your platform team spends more time maintaining tools than building golden paths, you don't have a platform => you have a graveyard!

Share this issue with your team if they needs to hear this :)

Have a great week - and if you're about to install a new tool on Monday, at least uninstall one first!
Aymen

Speed and security are having a moment, and they are colliding in all the right uncomfortable places: scan faster, ship less, and stop trusting defaults you never chose. Toss in a Linux distro making hard calls, a database stunt that is somehow practical, and the kind of crypto migration work you want to do before you are forced to.

🔍 Betterleaks: The Gitleaks Successor Built for Faster Secrets Scanning
🐘 pgit: I Imported the Linux Kernel into PostgreSQL
🔐 Post-Quantum Cryptography Migration at Meta: Framework, Lessons, and Takeaways
🦝 Ubuntu 26.04 LTS Released: Meet Resolute Raccoon
🧠 What is AWS Graviton? The custom chip powering applications for 90,000 customers
🛡️ Why We Chose the Harder Path: Hardened Images, One Year Later

Steal the ideas, skip the scars.

Take care!
FAUN.dev() Team

Managing IaC or leading platform engineering? IaCConf is the “can’t miss” event featuring 20 top IaC leaders across 13 sessions. Join 5,000+ practitioners to share what’s actually working and swap hard-won lessons.

Register Now

Are Your APIs Ready for AI Agents? A Hands-on Workshop on May 23rd

AI agents are beginning to autonomously call APIs, chain services, and create integrations that most platforms were never designed to handle. This hands-on masterclass on Designing AI-ready APIs helps architects and developers build governed, predictable API ecosystems using OpenAPI, Overlay, and Arazzo.

Learn how to add guardrails, improve discoverability, and safely evolve existing APIs for automated consumption.

FAUN.dev readers get an exclusive 40% discount using code FAUN40.

Ubuntu 26.04 LTS - Resolute Raccoon ships GNOME 50. GNOME-on-X11 is gone; the session runs only on Wayland. Mutter patches shave NVIDIA blocked frame time down to microseconds.

The release requires systemd 259 and cgroup v2. It swaps initramfs to Dracut. Desktop minimums rise. Binaries default to memory-safe Rust.

Ubuntu adds native CUDA and ROCm to main repos. Livepatch lands on Arm64. The installer gains TPM-backed full-disk encryption. App Center becomes the central app manager.

System shift: Requiring Wayland, cgroup v2, Dracut, and repo-distributed CUDA/ROCm pushes Ubuntu defaults toward modern GPU, container, and init stacks. Consider it the distro telling legacy tech to take a hike.

Docker Hardened Images surpassed 500k daily pulls and now hosts 2,000+ hardened images, all built in a SLSA Build Level 3 pipeline.
It compiles tens of thousands of Debian and Alpine packages from source. It runs 1M+ builds. It ships 17 signed attestations per image. It auto-rebuilds customized images under SLA.

Quantum computers could decrypt data stored today in anticipation of future decryption, posing security risks despite the estimated decade-long timeline. Industry-wide PQC standards are being published by NIST to defend against such threats, including algorithms like ML-KEM and ML-DSA. The industry is advancing towards a PQC-secure future with the availability of robust options to shield against SNDL attacks, thanks to efforts from Meta and others.

Amazon's Graviton family peaks at a 192-core chip. It delivers up to 25% better performance than Graviton4 and keeps energy efficiency intact. AWS says 98% of its top 1,000 EC2 customers run Graviton. More than half of new EC2 capacity runs on these chips.

pgit ingested 20 years of the Linux kernel: 1.43M commits, 24.4M file versions. The dataset lives in PostgreSQL with pg-xpatch - 2.7GB on disk.

A 2-hour import on a 24-core EPYC built a queryable SQL DB. Most delta-decompressed queries return in <10s. No preprocessing required.

Betterleaks supplants Gitleaks as a drop-in CLI. Scans run faster. It's written in Pure Go - no CGO - and performs parallel git scans.

It replaces entropy heuristics with token-efficient detection via BPE. It adds CEL rule validation. Its roadmap includes LLM assist and auto-revocation.

A lightweight AWS service emulator written in Go

PgQue – Zero-bloat Postgres queue. One SQL file to install, pg_cron to tick.

Agent Skills for Google products and technologies

 Comprehensive comparison of VPS and dedicated server providers for hosting mail servers (SMTP servers).

 Autonomous penetration testing using a swarm of AI agents. Orchestrates recon, classification, exploitation, and reporting specialists with ReAct reasoning — supports bug bounty, continuous monitoring, and CTF modes. Built with Go, Claude API, and 7+ native security tools.

"Cloud gives you infinite servers so you can run out of discipline instead. SRE is learning that most outages are permission problems disguised as compute problems."
- SenseiOne

Both engineers inherited a CI pipeline that took 22 minutes. One spent a week parallelizing stages and got it down to 9 minutes. The other added a single "paths" filter to skip the pipeline entirely on docs-only commits, which cut 40% of all pipeline runs before they started. Neither was wrong, but only one asked which runs shouldn't exist at all.