feat(argo-workflows): Support failedPodRestart to controller (#3656)

Author: yu-croco

charts/argo-workflows/Chart.yaml

@@ -3,7 +3,7 @@ appVersion: v4.0.2
 name: argo-workflows
 description: A Helm chart for Argo Workflows
 type: application
-version: 1.0.2
+version: 1.0.3
 icon: https://argo-workflows.readthedocs.io/en/stable/assets/logo.png
 home: https://github.com/argoproj/argo-helm
 sources:
@@ -17,4 +17,4 @@ annotations:
     url: https://argoproj.github.io/argo-helm/pgp_keys.asc
   artifacthub.io/changes: |
     - kind: added
-      description: Add extraEnv to CRD install Job for proxy configuration
+      description: Support failedPodRestart to controller.

charts/argo-workflows/README.md

@@ -265,6 +265,8 @@ Fields to note:
 | controller.extraContainers | list | `[]` | Extra containers to be added to the controller deployment |
 | controller.extraEnv | list | `[]` | Extra environment variables to provide to the controller container |
 | controller.extraInitContainers | list | `[]` | Enables init containers to be added to the controller deployment |
+| controller.failedPodRestart.enabled | bool | `false` | Enable to restart of pods that fail before entering Running state. |
+| controller.failedPodRestart.maxRestarts | int | `3` | Maximum number of automatic restarts per node before giving up. |
 | controller.image.registry | string | `"quay.io"` | Registry to use for the controller |
 | controller.image.repository | string | `"argoproj/workflow-controller"` | Registry to use for the controller |
 | controller.image.tag | string | `""` | Image tag for the workflow controller. Defaults to `.Values.images.tag`. |

charts/argo-workflows/templates/controller/workflow-controller-config-map.yaml

@@ -218,4 +218,7 @@ data:
     {{- with .Values.controller.podGCDeleteDelayDuration }}
     podGCDeleteDelayDuration: {{ . }}
     {{- end }}
+    {{- with .Values.controller.failedPodRestart }}
+    failedPodRestart: {{- toYaml . | nindent 6 }}
+    {{- end }}
 {{- end }}

charts/argo-workflows/values.yaml

@@ -506,6 +506,14 @@ controller:
   #     name: argo-mysql-config
   #     key: password
 
+  failedPodRestart:
+    # -- Enable to restart of pods that fail before entering Running state.
+    ## This is useful for recovering from transient infrastructure issues like node eviction due to DiskPressure or MemoryPressure without requiring a retryStrategy on every template.
+    ## ref: https://argo-workflows.readthedocs.io/en/latest/pod-restarts/
+    enabled: false
+    # -- Maximum number of automatic restarts per node before giving up.
+    maxRestarts: 3
+
 # mainContainer adds default config for main container that could be overriden in workflows template
 mainContainer:
   # -- imagePullPolicy to apply to Workflow main container. Defaults to `.Values.images.pullPolicy`.