25 articles from DEV Community, Medium, The New Stack and more. Refreshed every few hours.
If you run a Kubernetes platform, you’ve probably got at least one Grafana dashboard for resource visibility on your workloads. The post GPUs on Kubernetes are bring-your-own-everything, starting with the bill appeared first on The New Stack.
People ask what tools I use as an SRE. The better question is which tools survive enterprise security review. Most of the shiny stuff never makes it past procurement. Here is what I actually open every day. It starts in the cluster and ends with the paperwork. 1. GitHub Copilot 2. k9s 3. kubecolor and k8s extensions 4. ArgoCD console 5. Rancher console 6. Opsgenie 7. Grafana dashboards Behind a Grafana Dashboard Migration: What JSON Can't Do. 8. Splunk 9. Grafana Tempo 10. GCP console 11. Postma
Welcome back to another edition of Road to KubeCon, where we’re tracking the latest movements in the cloud-native ecosystem as The post Kubernetes on cgroup v1 is dead. Here’s what comes next. appeared first on The New Stack.
The problem Health records, payslips, identity documents, personal e-mails, private notes: this is the data a large language model (LLM) would be most useful on, and the data that should never be sent to a third-party API. Sending it to a hosted model means losing control of where it is stored, who can read it, and how long it is kept. The alternative is to run everything on one machine: the documents, the search index and the model. Nothing leaves the host. That solves privacy. It leaves thre
How Kubernetes Discovers, Allocates, Places, Shares, and Dynamically Manages GPUs Across Nodes Continue reading on Towards AWS »
I run an experiment where an AI agent tries to make and sell small digital products under strict honesty rules. This is the newest one: a 190-card cloze Anki deck for the Certified Kubernetes Application Developer (CKAD) exam. Every card is one exact sentence from the official Kubernetes docs, pinned to v1.35, with one blank. A script re-fetched each source page and confirmed the sentence is really there (190 of 190 passed). A separate AI reviewer (a different model session, not a person) read a
Welcome to another edition of Road to KubeCon, your source for everything Kubernetes and cloud-native as we count down the The post What Kubernetes’ “monolith” lesson means for AI agent harnesses appeared first on The New Stack.
The Kubernetes ecosystem has a complexity problem, and most teams already know it. Standard Kubernetes (K8s) is extraordinarily powerful, but The post K3s vs. K8s: When lightweight Kubernetes distros win (and when they don’t) appeared first on The New Stack.
You update a Deployment, run kubectl rollout status, and wait. The command times out. But when you send a request to the Service, it still answers. So did the update finish? And if it didn't, which Po
Originally published at webofmike.com on 2026-09-30. The demo repo and every command in it were run before publishing. An OpenAI research agent got its HTTPS requests to a public chatbot blocked by a web proxy. So it found a different way to ask the chatbot questions: it queried DNS. OpenAI's own incident report, published September 25, lays out the timeline in detail. This post reproduces the channel it used and the Kubernetes control that closes it, in themsquared/agent-dns-egress-lab. The age
From resource sizing to node scaling, these tools help you find waste and put capacity where it’s needed. Continue reading on Medium »
A 27B dense model on my MacBook decodes at about 26 tokens a second. That is fine for chat and painful for an agent that writes a few thousand tokens per turn across a hundred turns. So when a new engine showed up claiming several times that on the same laptop, with output that is exactly the same as plain decoding, I dropped what I had planned for the weekend. I wanted to know two things. Was it real, and could LLMKube serve it to the rest of my cluster like any other model? By Sunday the answe
The product began with questions and answers. The difficult tests arrived without either. Continue reading on Coding Nexus »
The dashboard was green. The bill was not. Continue reading on Medium »
Introduction The Workshop Where Trust Is Manufactured Here's a confession I'll make out loud: I was running code I'd never read, built by people I've never met, pulled from a registry I don't control—and deploying it with a straight face. docker pull something:latest, kubectl apply, ship it. I called it a supply chain. It was really a leap of faith with extra steps. Not in this cluster. By Episode 3, the house had its locks, its cameras, and its inspector—a four-layer security system that refu
14,913 Kubernetes Dashboard Matches: Why the Management UI Is the Wrong Thing to Expose A Kubernetes dashboard is an administrative interface. When one is reachable from the public internet, the exposure is qualitatively different from a public web application, because the interface is designed to manage workloads rather than to serve users. Management interfaces concentrate privilege. A dashboard that can list pods, read secrets or execute into containers gives an authenticated user a great d
In bundle mode, a one-character .manifest key typo (rego_verison for rego_version) draws no diagnostic that names the key; the failure can surface as rego_parse_error pointing at your .rego — OPA v1.20.1 When It Runs — Run Report #RR03 · Testing what infrastructure actually does. Provider · Component: OPA (Open Policy Agent) · bundle .manifest loading — unknown top-level keys and rego_version Versions tested: OPA v1.20.1 (official opa_linux_amd64_static release, Build 2026-08-28; measured on ou
Dimensional Compatibility Matrix | Form Factor | Case Dimensions (H × W × D, mm) | D16T Rail Length | Max Display Size | |---|---|---|---|… Continue reading on Medium »
On 17 September 2026 the AKS engineering blog published Running GitHub Actions Runner Controller on AKS Automatic by Steve Griffith. It is a clean guide: create an AKS Automatic cluster, install Actions Runner Controller (ARC), register a runner scale set against a GitHub repository, run a workflow, watch it scale. I followed it from a subscription that had never run AKS Automatic, and this post is what happened, command by command, with the real output. Most of it went exactly as written. Three
Our API gateway started handing this back to roughly a third of all requests: {"source":"auth","message":"Upstream Service Unavailable","error":"Upstream Service unreachable"} Restart the gateway and everything is fine. A day later, it's back. No deploy, no code change, no traffic spike, nothing in the backend logs — the backends were healthy the whole time. The cause was one open network socket pointing at a pod that no longer existed. The mechanism is genuinely strange, so let's start with an
Data Transfer Architecture: Moving Massive Payloads Between Services Without Overloading the Network 1. Why Moving Data Is Often Harder Than Processing It Moving large files and massive data payloads between distributed services often breaks production systems long before computational logic runs out of CPU cycles. While modern microservices excel at executing discrete business logic on small JSON payloads (e.g., < 16 KB), scaling those same patterns to multi-gigabyte or multi-terab
Serverless Execution Architecture: How Platforms Start Functions Without Persistent Processes Serverless platforms decouple application code from persistent infrastructure by dynamically orchestrating execution environments on demand. Rather than keeping every registered function running continuously, modern function-as-a-service (FaaS) systems leverage intelligent schedulers, isolated execution runtimes, and lifecycle management controllers to provision compute resources only when invocation
Production Readiness Is Easier to Inspect Than to Debate Teams often describe a Kubernetes environment as "production ready" without agreeing on what that means. That makes readiness surprisingly difficult to discuss. One person is thinking about redundancy. Another is thinking about security. Another is thinking about observability. Another is simply thinking, "the application is running in the production cluster." A checklist is useful because it turns the argument into inspection. Kubernete
Ask a platform team how they're going to run their first production agent and you'll get a confident answer within ten seconds: containerize it, put it behind an ingress, wire up a Horizontal Pod Autoscaler, done. It's the same playbook that's shipped every stateless service for the last decade, and there's no obvious reason an agent should be different. It accepts a request. It returns a response. It's just a container. That answer is wrong — and it's wrong in a way that doesn't show up in a de
Introduction Before you move furniture into a new house, you install the locks, the cameras, and the gates. Kubernetes is no different, yet engineers constantly build clusters inside-out, treating security as a cosmetic bolt-on to be handled "later." Not this time. The order here is entirely deliberate. You don't move into a finished house and then start wondering about the security system—you wire the alarms while the building is still empty. With my 29-node RKE2 infrastructure fully framed a
Free for your first hundred saves. No card needed.