From cbbe199633988109f880a71005c5b4e5a0c16514 Mon Sep 17 00:00:00 2001 From: "azykov@mail.ru" Date: Mon, 7 Sep 2026 09:46:09 +0300 Subject: [PATCH] longhorn generous readiness probe 5s timeout --- .../longhorn/longhorn-system/helmrelease.yaml | 22 +++++++++++++++++++ 1 file changed, 22 insertions(+) diff --git a/clusters/base/infra/longhorn/longhorn-system/helmrelease.yaml b/clusters/base/infra/longhorn/longhorn-system/helmrelease.yaml index 7c459ed..23f0659 100644 --- a/clusters/base/infra/longhorn/longhorn-system/helmrelease.yaml +++ b/clusters/base/infra/longhorn/longhorn-system/helmrelease.yaml @@ -24,6 +24,28 @@ spec: upgrade: remediation: retries: 3 + postRenderers: + - kustomize: + patches: + - target: + kind: DaemonSet + name: longhorn-manager + # chart hardcodes readinessProbe without timeoutSeconds, so it falls back to the + # k8s default of 1s; under CPU contention (e.g. busy nodes) that's too tight and + # causes spurious "Readiness probe failed ... Client.Timeout exceeded" flapping + patch: | + apiVersion: apps/v1 + kind: DaemonSet + metadata: + name: longhorn-manager + spec: + template: + spec: + containers: + - name: longhorn-manager + readinessProbe: + timeoutSeconds: 5 + failureThreshold: 5 values: defaultSettings: defaultReplicaCount: 3 # по числу нод control-plane