restart the container on failure

kubeflow · Mar 14, 2023 · 214a736 · 214a736
1 parent 7067269
commit 214a736
Showing 1 changed file with 4 additions and 4 deletions.
diff --git a/sdk/python/test/e2e/test_e2e_xgboostjob.py b/sdk/python/test/e2e/test_e2e_xgboostjob.py
@@ -51,7 +51,7 @@ def test_sdk_e2e_with_gang_scheduling():
 
     master = V1ReplicaSpec(
         replicas=1,
-        restart_policy="Never",
+        restart_policy="OnFailure",
         template=V1PodTemplateSpec(spec=V1PodSpec(
             containers=[container],
             scheduler_name=get_pod_spec_scheduler_name(GANG_SCHEDULER_NAME),
@@ -60,7 +60,7 @@ def test_sdk_e2e_with_gang_scheduling():
 
     worker = V1ReplicaSpec(
         replicas=1,
-        restart_policy="Never",
+        restart_policy="OnFailure",
         template=V1PodTemplateSpec(spec=V1PodSpec(
             containers=[container],
             scheduler_name=get_pod_spec_scheduler_name(GANG_SCHEDULER_NAME),
@@ -104,13 +104,13 @@ def test_sdk_e2e():
 
     master = V1ReplicaSpec(
         replicas=1,
-        restart_policy="Never",
+        restart_policy="OnFailure",
         template=V1PodTemplateSpec(spec=V1PodSpec(containers=[container])),
     )
 
     worker = V1ReplicaSpec(
         replicas=1,
-        restart_policy="Never",
+        restart_policy="OnFailure",
         template=V1PodTemplateSpec(spec=V1PodSpec(containers=[container])),
     )