Devops

Picking a log aggregator in 2026: Loki for cheap storage, ClickHouse for query power, OpenSearch for full-text, Datadog when you can pay. Decision matrix and patterns.

Practical Kubernetes cost reduction in 2026. Right-size requests / limits, Karpenter for node management, VPA / HPA, spot pools, and the dashboards that find waste.

Modern Dockerfile patterns: multi-stage builds, BuildKit cache mounts, distroless images, non-root users, and the boring habits that produce small, fast, secure images.

Chaos engineering done right. Game days, failure injection (Chaos Mesh, Gremlin), what to test, the observability needed, and the cultural shifts that make it stick.

Picking a workflow orchestrator in 2026. Argo Workflows for Kubernetes-native; Airflow for mature ETL; Dagster for data-aware orchestration; Prefect for Python-first. The decision matrix.

Tactics that produce real cloud savings: right-sizing, spot, savings plans, egress, idle cleanup, multi-cloud arbitrage, FinOps culture, and the diagnostic tools that find waste.

An honest 2026 take on serverless. The use cases that won (event-driven, edge, low-traffic APIs), the ones that didn’t (long-running, stateful), and the operational realities.

What changed in observability since 2020. Wide events vs three-pillars, SLOs as the unit of conversation, OTel’s role, and how to actually find problems in production.

What good CI/CD looks like in 2026. Pipeline speed, caching, parallelism, branch protection, deploy gates, and the boring habits that compound.

How to manage secrets in production in 2026. Vault, ESO, cloud secret managers, sealed secrets, rotation, audit, and the developer-friendly patterns.