[{"data":1,"prerenderedAt":16},["ShallowReactive",2],{"blog-tag-ja-prometheus-adapter":3},[4],{"path":5,"title":6,"description":7,"date":8,"tags":9},"\u002Fblog\u002Fja\u002Fk3s-hpa-prometheus-adapter-discovery-interval-stale-metrics","discoveryIntervalを疑え。K3sのHPAがPrometheus Adapter経由で\"古いメトリクス\"を掴み続ける設定ミス","K3sのHPAがCPU使用率は正常なのにスケールしない。原因はPrometheus Adapterのmetrics-relist-interval\u002Fmetrics-max-ageとHPA sync-periodが噛み合っていない設定ミスだった。3層の遅延構造と診断コマンドを解説する。","2026-10-12",[10,11,12,13,14,15],"k3s","kubernetes","hpa","prometheus-adapter","autoscaling","observability",1792592674137]