New in #TalosLinux & Omni
→ Multi-doc configs for networking
→ Talos Linux cluster imports for Omni
→ Pre-seeded images
→ Built-in OOM handling
#platformengineering
Ramp revealed the architecture of Inspect!
Internal #AIagents are now handling ~30% of merged pull requests across their frontend and backend repositories.
How❓ By giving #AI the same access to the development environment as human engineers.
Check out #InfoQ for a deep dive: https://bit.ly/4btmhBS
Speaker Reveal 🚀
Using multiple event brokers? That often means scattered events and poor discoverability.
In this session by Prachi Jamdade, see how Gravitee unifies event streams into a single, searchable AsyncAPI catalog with built-in governance to improve quality, security, and consistency.
Grab a free open pass and join this session.
Agenda and Free Passes: https://conference.asyncapi.com/venue/California
#EventDrivenArchitecture #AsyncAPI #EnterpriseAPIs #PlatformEngineering
Join me next Thursday (February 5) at Cloud Platform Engineering London.
I'll be talking about our Platform Engineering research and you can also hear about cloud platforms from Sean Tracey and platform planes from Immo Hüneke.
TechTalk #11 is here
We're kicking off the year by wrapping up our PaaS series. The focus: #OpenTelemetry 🔭
Standardizing on #OTel is a win for developers, but it creates a massive #Observability data challenge for the #PlatformEngineering team.
🎙️ Join Mathias Palmersheim and Marc Sherwood as we discuss:
The apply before-or-after merge debate is perennial, but I find this article from Terramate covers it thoroughly and has a nice tabulated summary at the end.
#Terramate #Terraform #Terragrunt #DevOps #OpenTofu #IaC #Cloud #PlatformEngineering
Most Posit issues do not start with a visible failure.
They build up through configuration choices, version drift, and infrastructure decisions that no longer fit how the platform is used.
Over time, this leads to unstable deployments, timeouts, security concerns, and harder upgrades.
At Jumping Rivers, our Posit Health Check reviews configuration, servers, versions, and security to surface risk early.
Contact us at hello@jumpingrivers.com
CFP closes soon: DevOpsDays Zürich 2026
Deadline: 31 Jan 2026, 23:59 (CET)
We’re looking for:
• Keynotes (45m)
• Talks (25m)
• Ignite (5m)
• Workshops (3h)
Submit: https://buff.ly/WRIAYeB
Event: 06–07 May 2026, Zürich
What topic do you want to see this year?
Ended up kicking Docker Hardened Images to the curb.
The good folks at #Docker decided to update their #Redis image to use a distro-less start (no shell, no other utilities) *without changing the image tag*.
So when I started up my #Mastodon server after an upgrade, my customization script wouldn’t run in the container and I had to scramble to find an alternative.
Basically, they broke the contract of immutable image version tags. Won’t trust them again, even if it did harden #security further.
So this server is back to a good ol’ upstream #PostgreSQL image on #AlpineLinux. And #Valkey on #Alpine, too, because Redis Ltd..’s software licensing shenanigans are sketchy.
Shane is worried about backups for his janky Kubernetes homelab. The rest of us advise him on exactly what to back up, how to go about picking an offsite backup location and setting it up, how best to back up databases, and more.
In 2026, the best Software Architectures aren’t just technical - they’re Sociotechnical.
It’s about:
🔹 Platforms & automation at scale
🔹 Decentralized decision-making
🔹 Architects as coaches & enablers
📥 #FreeDownload the #InfoQ eMag and learn how to build resilient, scalable & observable architectures for AI-driven environments 👉 https://bit.ly/4qFFIfs
Part of the ongoing learning that there are teams getting value from AI agents right now but (particularly when scale is involved), they're spending their time building platforms that let other teams deploy agents safely. The agent itself is almost the easy part. The hard part is the registry, the router, the queue, the audit trail that surrounds it, and so on.
Links to the two case studies in the thread ⬇️
We’re excited to welcome KPMG as a sponsor of DevOpsDays Zürich 2026! 🙌
KPMG helps organizations build trust, manage risk, and deliver complex transformations. The everyday reality of modern DevOps.
With their support, we keep DevOpsDays community-first:
✅ talks & ignites
✅ open spaces
✅ 2 days of learning + networking
#DevOpsDays #DevOpsDaysZurich #DevOps #PlatformEngineering #SRE #Cloud #DataAI #Community #Zurich #Winterthur
A practical guide to ensuring high availability in Kubernetes with ReplicationControllers. Covers pod replication, fault tolerance, self-healing apps, and production-grade Kubernetes design.
🔗 https://shorturl.at/605FM
#Kubernetes #K8s #HighAvailability #CloudNative #DevOps #aws #SRE #Containers #PlatformEngineering
I hadn't thought much about this until someone asked whether ZenML locks you into Python-only codebases. The answer is more nuanced than I expected.
Put together an example repo with a multi-stage Dockerfile for remote orchestration and some sample earnings transcripts. Still want to try adding Rayon for parallel processing within a single step.
Example repo: https://github.com/zenml-io/rust-pipeline/
Engineering GenAI solutions on AWS is moving fast and solid architecture matters.
We’re hosting a practical session for solution, cloud, AI, and platform architects, focused on how these systems are actually designed and deployed on AWS.
🎟️ Register here:
https://www.eventbrite.com/e/engineering-genai-solutions-on-aws-tickets-1979979561167?aff=Ebieestack
🔥 40% off with code: FIRST40 (limited)
Speakers include architects from AWS sharing real‑world perspectives.
#AWS #GenerativeAI #CloudArchitecture #DevRel #PlatformEngineering
Running Kubernetes isn’t enough anymore.
If you can’t see your systems - you don’t control them. 👀
Observability just caught up with Kubernetes.
That’s the real shift.
My new short: https://youtube.com/shorts/TwKlvGVEXC8
#DevOps #Observability #Kubernetes #CloudNative #PlatformEngineering
Hi, I’m Mark. Glad you’re here.
I work in #PlatformEngineering and #SoftwareEngineering, after a long stretch in #Perl and now deep into #Go, #Docker, #Kubernetes, and #AWS. I like systems that behave, tools that justify their existence, and explanations that don’t waste anyone’s time.
https://PhoenixTrap.com is where I #blog about the engineering side of that: production, clarity, and the discipline required to build things that hold up under pressure.
I’m also an #Objectivist, which for me means treating ideas as tools for living, not ornaments.
https://TheBeautifulPrison.com is where I explore — through #fiction and #essay #writing — the philosophical side: consciousness, identity, boundaries, and what happens when a self is tested. If you’re curious about the deeper themes behind my posts, that’s the place to look.
Outside of work and writing, you’ll see #BassGuitar, conventions like #DragonCon, and the occasional bit of structured nonsense.
This #Mastodon instance is #SelfHosted because I like owning my infrastructure and my identity; it keeps the whole thing honest.
If something I say sparks a thought, feel free to reply. I enjoy good conversations — especially the kind where two people are actually trying to understand each other.
Pull up a chair.
What stood out to me: the enemy isn't cron syntax. It's synchronization. "Every hour" isn't a schedule—it's a traffic pattern you're committing to. And at scale, human bias toward readable times becomes a reliability problem.
The harder question: do you enforce this at the platform level, or trust engineers to pick good times? Shopify chose enforcement.
Has anyone else implemented schedule jitter to save their clusters?
Client Info
Server: https://mastodon.social
Version: 2025.07
Repository: https://github.com/cyevgeniy/lmst