Review remarks

cluster-autoscaler/config/autoscaling_options.go

@@ -300,7 +300,7 @@ type AutoscalingOptions struct {
 	// ProvisioningRequestMaxBackoffTime is the max time for ProvisioningRequest be considered by CA after failed ScaleUp request.
 	ProvisioningRequestMaxBackoffTime time.Duration
 	// ProvisioningRequestMaxCacheSize is the max size for ProvisioningRequest cache that is stored for retry backoff.
-	ProvisioningRequestMaxCacheSize int
+	ProvisioningRequestMaxBackoffCacheSize int
 }
 
 // KubeClientOptions specify options for kube client

cluster-autoscaler/main.go

@@ -269,14 +269,14 @@ var (
 			"--max-graceful-termination-sec flag should not be set when this flag is set. Not setting this flag will use unordered evictor by default."+
 			"Priority evictor reuses the concepts of drain logic in kubelet(https://github.com/kubernetes/enhancements/tree/master/keps/sig-node/2712-pod-priority-based-graceful-node-shutdown#migration-from-the-node-graceful-shutdown-feature)."+
 			"Eg. flag usage:  '10000:20,1000:100,0:60'")
-	provisioningRequestsEnabled           = flag.Bool("enable-provisioning-requests", false, "Whether the clusterautoscaler will be handling the ProvisioningRequest CRs.")
-	provisioningRequestInitialBackoffTime = flag.Duration("provisioning-request-initial-backoff-time", 1*time.Minute, "Initial backoff time for ProvisioningRequest retry after failed ScaleUp.")
-	provisioningRequestMaxBackoffTime     = flag.Duration("provisioning-request-max-backoff-time", 10*time.Minute, "Max backoff time for ProvisioningRequest retry after failed ScaleUp.")
-	provisioningRequestMaxCacheSize       = flag.Int("provisioning-request-max-cache-size", 1000, "Max size for ProvisioningRequest cache size used for retry backoff mechanism.")
-	frequentLoopsEnabled                  = flag.Bool("frequent-loops-enabled", false, "Whether clusterautoscaler triggers new iterations more frequently when it's needed")
-	asyncNodeGroupsEnabled                = flag.Bool("async-node-groups", false, "Whether clusterautoscaler creates and deletes node groups asynchronously. Experimental: requires cloud provider supporting async node group operations, enable at your own risk.")
-	proactiveScaleupEnabled               = flag.Bool("enable-proactive-scaleup", false, "Whether to enable/disable proactive scale-ups, defaults to false")
-	podInjectionLimit                     = flag.Int("pod-injection-limit", 5000, "Limits total number of pods while injecting fake pods. If unschedulable pods already exceeds the limit, pod injection is disabled but pods are not truncated.")
+	provisioningRequestsEnabled            = flag.Bool("enable-provisioning-requests", false, "Whether the clusterautoscaler will be handling the ProvisioningRequest CRs.")
+	provisioningRequestInitialBackoffTime  = flag.Duration("provisioning-request-initial-backoff-time", 1*time.Minute, "Initial backoff time for ProvisioningRequest retry after failed ScaleUp.")
+	provisioningRequestMaxBackoffTime      = flag.Duration("provisioning-request-max-backoff-time", 10*time.Minute, "Max backoff time for ProvisioningRequest retry after failed ScaleUp.")
+	provisioningRequestMaxBackoffCacheSize = flag.Int("provisioning-request-max-backoff-cache-size", 1000, "Max size for ProvisioningRequest cache size used for retry backoff mechanism.")
+	frequentLoopsEnabled                   = flag.Bool("frequent-loops-enabled", false, "Whether clusterautoscaler triggers new iterations more frequently when it's needed")
+	asyncNodeGroupsEnabled                 = flag.Bool("async-node-groups", false, "Whether clusterautoscaler creates and deletes node groups asynchronously. Experimental: requires cloud provider supporting async node group operations, enable at your own risk.")
+	proactiveScaleupEnabled                = flag.Bool("enable-proactive-scaleup", false, "Whether to enable/disable proactive scale-ups, defaults to false")
+	podInjectionLimit                      = flag.Int("pod-injection-limit", 5000, "Limits total number of pods while injecting fake pods. If unschedulable pods already exceeds the limit, pod injection is disabled but pods are not truncated.")
 )
 
 func isFlagPassed(name string) bool {
@@ -451,7 +451,7 @@ func createAutoscalingOptions() config.AutoscalingOptions {
 		AsyncNodeGroupsEnabled:                  *asyncNodeGroupsEnabled,
 		ProvisioningRequestInitialBackoffTime:   *provisioningRequestInitialBackoffTime,
 		ProvisioningRequestMaxBackoffTime:       *provisioningRequestMaxBackoffTime,
-		ProvisioningRequestMaxCacheSize:         *provisioningRequestMaxCacheSize,
+		ProvisioningRequestMaxBackoffCacheSize:  *provisioningRequestMaxBackoffCacheSize,
 	}
 }
 
@@ -530,7 +530,7 @@ func buildAutoscaler(debuggingSnapshotter debuggingsnapshot.DebuggingSnapshotter
 			return nil, err
 		}
 		opts.LoopStartNotifier = loopstart.NewObserversList([]loopstart.Observer{provreqProcesor})
-		injector, err := provreq.NewProvisioningRequestPodsInjector(restConfig, opts.ProvisioningRequestInitialBackoffTime, opts.ProvisioningRequestMaxBackoffTime, opts.ProvisioningRequestMaxCacheSize)
+		injector, err := provreq.NewProvisioningRequestPodsInjector(restConfig, opts.ProvisioningRequestInitialBackoffTime, opts.ProvisioningRequestMaxBackoffTime, opts.ProvisioningRequestMaxBackoffCacheSize)
 		if err != nil {
 			return nil, err
 		}

cluster-autoscaler/processors/provreq/injector.go

@@ -101,7 +101,7 @@ func (p *ProvisioningRequestPodsInjector) GetPodsFromNextRequest(
 	}
 	for _, pr := range provReqs {
 		if !isSupportedClass(pr) {
-			klog.Warningf("Provisioning Class %s is not supported", pr.Spec.ProvisioningClassName)
+			klog.Warningf("Provisioning Class %s is not supported for ProvReq %s/%s", pr.Spec.ProvisioningClassName, pr.Namespace, pr.Name)
 			continue
 		}
 		conditions := pr.Status.Conditions