Fix set scheduler error (#1448)

* [fix]fix set scheduler error Signed-off-by: bert.li <[email protected]> * refactor gangScheduler set Signed-off-by: bert.li <[email protected]>
kubeflow · Nov 4, 2021 · 1a69a5e · 1a69a5e
1 parent 72e4387
commit 1a69a5e
Show file tree

Hide file tree

Showing 2 changed files with 12 additions and 3 deletions.
diff --git a/pkg/common/util/util.go b/pkg/common/util/util.go
@@ -50,3 +50,11 @@ func GetReplicaTypes(specs map[commonv1.ReplicaType]*commonv1.ReplicaSpec) []com
 	}
 	return keys
 }
+func GetSchedulerName(replicas map[commonv1.ReplicaType]*commonv1.ReplicaSpec) string {
+	for _, spec := range replicas {
+		if len(spec.Template.Spec.SchedulerName) > 0 {
+			return spec.Template.Spec.SchedulerName
+		}
+	}
+	return ""
+}
diff --git a/pkg/controller.v1/tensorflow/tfjob_controller.go b/pkg/controller.v1/tensorflow/tfjob_controller.go
@@ -829,12 +829,13 @@ func (r *TFJobReconciler) createNewPod(tfjob *tfv1.TFJob, rt, index string, spec
 	// 1. if user has specified other scheduler, we report a warning without overriding any fields.
 	// 2. if no SchedulerName is set for pods, then we set the SchedulerName to "volcano".
 	if r.Config.EnableGangScheduling {
-		if !util.IsGangSchedulerSet(replicas, gangSchedulerName) {
+		podSchedulerName := util.GetSchedulerName(replicas)
+		if len(podSchedulerName) == 0 {
+			podTemplate.Spec.SchedulerName = gangSchedulerName
+		} else if strings.Compare(podSchedulerName, gangSchedulerName) != 0 {
 			errMsg := "Another scheduler is specified when gang-scheduling is enabled and it will not be overwritten"
 			logger.Warning(errMsg)
 			r.Recorder.Event(tfjob, v1.EventTypeWarning, podTemplateSchedulerNameReason, errMsg)
-		} else {
-			podTemplate.Spec.SchedulerName = gangSchedulerName
 		}
 
 		if podTemplate.Annotations == nil {