Nick Nikolakakis

--------------------------

role: Principal SRE, Platform & AI Engineer
focus: Kubernetes · CNCF · Platform · AI
building: AI agents · LLMOps · MCP
shell: zsh
since: 2011

// The SRE Diary: Chronicles of Cloud, Kubernetes and AI

cat latest ./whoami git clone

$ ls -la ~/posts 55 entries

~/posts/auditing-proxies-and-meshes-for-the-kubectl-websockets-transport 21 min · 4468 words

$ Before You Upgrade: Auditing Your Proxies and Meshes for the kubectl WebSockets Transport

// Kubernetes is retiring the SPDY streaming transport behind kubectl exec, attach, cp and port-forward in favour of WebSockets. This is an audit you can run against your clusters this week to find the hop that will break, before the upgrade finds it for you.

~/posts/immutable-kubernetes-nodes-as-oci-images-bootc 21 min · 4409 words

$ Immutable Kubernetes Nodes as OCI Images - A Close Look at bootc

// What actually changes in how you build, roll out and roll back a Kubernetes node when the node OS ships as an OCI image instead of a package set, using bootc and a small worked example called Corium.

~/posts/kubernetes-gang-scheduling-workload-api 17 min · 3485 words

$ Kubernetes 1.37 Gang Scheduling - The Workload API, a Silently Dropped Field, and Where Preemption Stops

// KEP-4671 landed gang scheduling in Kubernetes 1.37 beta, behind a feature gate that is off by default. A four-node Kind cluster with six fake DRA GPUs shows the partial-placement deadlock, the single field that fixes it, and the two limits nobody documents: the Job field is pruned silently without a second alpha gate, and workload-aware preemption cannot reclaim DRA devices.

~/posts/modelplane-beyond-placement-routing-dynamo-gotchas 22 min · 4548 words

$ Modelplane Beyond Placement - Routing Policy, the Dynamo Stack, and Four Gotchas

// Part two of the Modelplane fleet inference series, covering the layer above the scheduler. ModelService is a routing policy rather than a load balancer, spec.stack Dynamo swaps the whole gang-scheduling machinery, and four documented gotchas will cost you an afternoon each.

~/posts/modelplane-fleet-inference 29 min · 5971 words

$ Modelplane - Fleet Inference Across Clusters, Demoed Without a Single GPU

// A hands-on tour of Modelplane, the Crossplane-based control plane for AI inference across a fleet of GPU clusters. Two regions on Kind with fake DRA GPUs, the platform/developer contract written in DRA's own vocabulary, fleet-level drain, and the release trap that costs you two APIs.

~/posts/agent-sandbox 19 min · 3912 words

$ Agent Sandbox - Kubernetes-Native Sandboxes for AI Agent Code Execution

// A hands-on tour of agent-sandbox, the Kubernetes SIG Apps project for isolated, stateful, singleton workloads built for AI agent runtimes. Includes a Go agent with sandboxed tools on Kind, warm pools claiming in 53ms, Flux GitOps, and the gotchas the docs don't mention.

~/posts/llm-d-distributed-inference 21 min · 4271 words

$ llm-d - Kubernetes-Native Distributed LLM Inference at Scale

// A hands-on tour of llm-d, the CNCF Sandbox framework for distributed LLM inference on Kubernetes - inference-aware routing, prefill/decode disaggregation, and KV-cache offload. Includes a GPU-free demo on Kind using the vLLM simulator, wired with Flux GitOps.

~/posts/choragos-multi-agent-orchestrator 26 min · 5350 words

$ Choragos - Running a Team of Coding Agents Without Losing Control

// Choragos runs a team of AI coding agents with a real division of labour: an orchestrator that plans and delegates, workers with their own context, model and credentials, and a delegate/work-done protocol carrying work between them. A walkthrough of v0.11.2 with config recipes for real teams.

~/posts/shipping-an-obsidian-plugin-with-a-coding-agent 9 min · 1825 words

$ Pinax - Shipping a Production-Grade Obsidian Plugin with a Coding Agent

// How I built and shipped Pinax, an LLM-buildable dashboard framework for Obsidian, using Claude Code: the guardrails that made agentic development work, the automated review battles, and why machine-checkable success criteria beat vibe coding.

~/posts/introducing-sphragis-eu-ai-act-compliance-gateway 17 min · 3410 words

$ Sphragis - The EU AI Act Compliance Gateway You Actually Control

// Sphragis is a self-hosted Go gateway that strips PII out of every LLM request and response before it leaves your network and writes a tamper-evident, hash-chained audit log. A walkthrough of the v0.3.0 release: local redaction, reversible tokenization, multi-provider routing, and OpenTimestamps anchoring.