This is an automated email from the ASF dual-hosted git repository.
adoroszlai pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/ozone-helm-charts.git
The following commit(s) were added to refs/heads/main by this push:
new 4017dff HDDS-14349. Adding startupProbe for HA deployments (#42)
4017dff is described below
commit 4017dff91a9cba309dc825f36513dd71e354469f
Author: Mark Tsai <[email protected]>
AuthorDate: Sat Sep 26 01:00:05 2026 +0800
HDDS-14349. Adding startupProbe for HA deployments (#42)
---
charts/ozone/templates/om/om-statefulset.yaml | 13 ++++++++++++-
charts/ozone/templates/scm/scm-statefulset.yaml | 15 ++++++++++++++-
charts/ozone/values.yaml | 18 ++++++++++++++++++
3 files changed, 44 insertions(+), 2 deletions(-)
diff --git a/charts/ozone/templates/om/om-statefulset.yaml
b/charts/ozone/templates/om/om-statefulset.yaml
index ac59c97..f9bc2f8 100644
--- a/charts/ozone/templates/om/om-statefulset.yaml
+++ b/charts/ozone/templates/om/om-statefulset.yaml
@@ -51,7 +51,7 @@ spec:
app.kubernetes.io/component: om
spec:
{{- if gt (int .Values.om.replicas) 1 }}
- terminationGracePeriodSeconds: 150
+ terminationGracePeriodSeconds: {{
.Values.om.terminationGracePeriodSeconds }}
{{- end }}
{{- with .Values.imagePullSecrets }}
imagePullSecrets: {{- toYaml . | nindent 8 }}
@@ -119,6 +119,17 @@ spec:
ozone admin om transfer -id {{ .Values.clusterId }} -r
|| true
sleep 10
fi
+ startupProbe:
+ exec:
+ command:
+ - sh
+ - -c
+ - |
+ ozone admin om roles -id {{ .Values.clusterId }} 2>/dev/null
| grep -q $(hostname) && exit 0 || exit 1
+ initialDelaySeconds: {{
.Values.om.startupProbe.initialDelaySeconds }}
+ periodSeconds: {{ .Values.om.startupProbe.periodSeconds }}
+ failureThreshold: {{ .Values.om.startupProbe.failureThreshold }}
+ timeoutSeconds: {{ .Values.om.startupProbe.timeoutSeconds }}
readinessProbe:
exec:
command:
diff --git a/charts/ozone/templates/scm/scm-statefulset.yaml
b/charts/ozone/templates/scm/scm-statefulset.yaml
index ca155f6..4a1c4a2 100644
--- a/charts/ozone/templates/scm/scm-statefulset.yaml
+++ b/charts/ozone/templates/scm/scm-statefulset.yaml
@@ -50,7 +50,7 @@ spec:
app.kubernetes.io/component: scm
spec:
{{- if gt (int .Values.scm.replicas) 1 }}
- terminationGracePeriodSeconds: 150
+ terminationGracePeriodSeconds: {{
.Values.scm.terminationGracePeriodSeconds }}
{{- end }}
{{- with .Values.imagePullSecrets }}
imagePullSecrets: {{- toYaml . | nindent 8 }}
@@ -137,6 +137,19 @@ spec:
ozone admin scm transfer -id {{ .Values.clusterId }} -r
|| true
sleep 10
fi
+ startupProbe:
+ exec:
+ command:
+ - sh
+ - -c
+ - |
+ ozone admin scm roles 2>/dev/null | grep -q $(hostname -f) \
+ && ozone admin safemode status 2>/dev/null | grep -q 'out of
safe mode' \
+ && exit 0 || exit 1
+ initialDelaySeconds: {{
.Values.scm.startupProbe.initialDelaySeconds }}
+ periodSeconds: {{ .Values.scm.startupProbe.periodSeconds }}
+ failureThreshold: {{ .Values.scm.startupProbe.failureThreshold }}
+ timeoutSeconds: {{ .Values.scm.startupProbe.timeoutSeconds }}
readinessProbe:
exec:
command:
diff --git a/charts/ozone/values.yaml b/charts/ozone/values.yaml
index 7d53b96..a01f249 100644
--- a/charts/ozone/values.yaml
+++ b/charts/ozone/values.yaml
@@ -152,6 +152,15 @@ om:
periodSeconds: 10
timeoutSeconds: 20
failureThreshold: 3
+ # Graceful termination period for Ozone Manager HA (om.replicas > 1),
+ # allowing leader transfer to complete before the pod is killed
+ terminationGracePeriodSeconds: 150
+ # Startup probe for Ozone Manager HA (used when om.replicas > 1)
+ startupProbe:
+ initialDelaySeconds: 30
+ periodSeconds: 10
+ timeoutSeconds: 20
+ failureThreshold: 30
# Ozone Manager service configuration
service:
type: ClusterIP
@@ -214,6 +223,15 @@ scm:
periodSeconds: 10
timeoutSeconds: 20
failureThreshold: 3
+ # Graceful termination period for Storage Container Manager HA (scm.replicas
> 1),
+ # allowing leader transfer to complete before the pod is killed
+ terminationGracePeriodSeconds: 150
+ # Startup probe for Storage Container Manager HA (used when scm.replicas > 1)
+ startupProbe:
+ initialDelaySeconds: 20
+ periodSeconds: 10
+ timeoutSeconds: 20
+ failureThreshold: 30
# Storage Container Manager service configuration
service:
type: ClusterIP
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
