observabilidad archivos

Container Monitoring: Beyond cAdvisor

cAdvisor was the historical default but today isn’t enough. How to combine eBPF, Kubernetes metrics, and APM for real container observability.

Fluent Bit replaces heavy agents with a C binary barely over a megabyte. When it beats Promtail or Vector and how to integrate it without drama.

SLOs only work if error budget is genuinely managed. How to define without ceremony and how to use them to balance speed and reliability.

Loki scales well until it doesn’t. How to label-index with judgement, separate ingestion paths, and avoid cardinality explosions in production.

Falco watches the kernel and detects anomalous behaviour without instrumenting apps. When it makes sense, how to integrate it, where ops hurts.

eBPF lets you run safe code inside the kernel for tracing, networking, and security. Why it’s the foundation of modern observability tooling.

PostgreSQL 16 brings logical replication from standby, better parallelism, and observability. The new features that really matter in production.

Grafana Labs offers a complete stack with Loki (logs), Tempo (traces), and Mimir (metrics). Technical analysis and alternatives comparison.

OpenTelemetry consolidates the three observability signals under a single standard. 2023 state, integration, and adoption path.

K8s 1.28 introduces native sidecars in alpha. What changes, how it affects service mesh and observability, and how to prepare for GA.