Azure AKS: diagnose HPA oscillation before raising replica limits
A production runbook for separating noisy metrics, miscalibrated requests, slow startup and node capacity before changing Horizontal Pod Autoscaler bounds on AKS.
Read article
Tag
prometheus
3 articles connected to this technical signal.
Azure Monitor Managed Prometheus: contain a cardinality spike before raising quotas
A production runbook for isolating the metric, target and labels behind an AKS time-series spike before filtering collection, adding capacity or rolling back.
Read articleOpenTelemetry Collector: diagnose telemetry drops before adding memory
A production runbook for locating telemetry loss across receivers, processors, queues and exporters, then validating capacity changes or rolling back safely.
Read article