Merge pull request #886 from kubernetes-sigs/master

Merge from master to release-0.5 branch

CHANGLOG.md

@@ -1,3 +1,27 @@
+## v0.5  
+
+**Closed issues:**
+
+- master and worker start in order in a job [\#872](https://github.com/kubernetes-sigs/kube-batch/issues/872)
+-  use kubernetes default schedule policies [\#864](https://github.com/kubernetes-sigs/kube-batch/issues/864)
+- v0.4 runtime panic  [\#861](https://github.com/kubernetes-sigs/kube-batch/issues/861)
+- Helm install fails [\#857](https://github.com/kubernetes-sigs/kube-batch/issues/857)
+- "NewResource/Convert2K8sResource" behavior mismatch [\#851](https://github.com/kubernetes-sigs/kube-batch/issues/851)
+- Phase of podgroup status looks incorrect [\#846](https://github.com/kubernetes-sigs/kube-batch/issues/846)
+- Update API GroupName [\#815](https://github.com/kubernetes-sigs/kube-batch/issues/815)
+- Migrate nodeorder and predicates plugins [\#814](https://github.com/kubernetes-sigs/kube-batch/issues/814)
+- Add Configuration for predicate plugin to enable/disable predicates algorithm [\#802](https://github.com/kubernetes-sigs/kube-batch/issues/802)
+- Consider support multi-containers pod error code handling [\#776](https://github.com/kubernetes-sigs/kube-batch/issues/776)
+- Deserved attr is not correctly calculated in proportion plugin [\#729](https://github.com/kubernetes-sigs/kube-batch/issues/729)
+- Keep backward compatibility for priority class [\#724](https://github.com/kubernetes-sigs/kube-batch/issues/724)
+- Change return value of NodeOrderFn from int to float [\#708](https://github.com/kubernetes-sigs/kube-batch/issues/708)
+- Add type Argument with some common parse function [\#704](https://github.com/kubernetes-sigs/kube-batch/issues/704)
+- Replace NodeOrder with BestNode [\#699](https://github.com/kubernetes-sigs/kube-batch/issues/699)
+- Support set default value to the configuration [\#695](https://github.com/kubernetes-sigs/kube-batch/issues/695)
+- Add resource predicates for tasks [\#694](https://github.com/kubernetes-sigs/kube-batch/issues/694)
+- Pass conformance test [\#589](https://github.com/kubernetes-sigs/kube-batch/issues/589)
+- big PodGroup blocks scheduling issue [\#514](https://github.com/kubernetes-sigs/kube-batch/issues/514)
+
 ## v0.4.2
 
 ### Notes:

Makefile

@@ -1,5 +1,5 @@
 BIN_DIR=_output/bin
-RELEASE_VER=v0.4.2
+RELEASE_VER=v0.5
 REPO_PATH=github.com/kubernetes-sigs/kube-batch
 GitSHA=`git rev-parse HEAD`
 Date=`date "+%Y-%m-%d %H:%M:%S"`

deployment/kube-batch/values.yaml

@@ -4,7 +4,7 @@
 replicaCount: 1
 image:
   repository: kubesigs
-  tag: v0.4.2
+  tag: v0.5
   pullPolicy: IfNotPresent
 resources:
   limits:

doc/usage/tutorial.md

@@ -14,7 +14,7 @@ To run `kube-batch`, a Kubernetes cluster must start up. Here is a document on [
 ```bash
 # mkdir -p $GOPATH/src/github.com/kubernetes-sigs/
 # cd $GOPATH/src/github.com/kubernetes-sigs/
-# git clone http://github.com/kubernetes-sigs/kube-batch -b v0.4.2
+# git clone http://github.com/kubernetes-sigs/kube-batch -b v0.5
 ```
 
 ### Deploy kube-batch by Helm
@@ -164,7 +164,6 @@ apiVersion: scheduling.k8s.io/v1beta1
 kind: PriorityClass
 metadata:
   name: high-priority
-  namespace: batch-ns01
 value: 1000
 ```
 

pkg/scheduler/api/node_info.go

@@ -158,6 +158,18 @@ func (ni *NodeInfo) SetNode(node *v1.Node) {
 	}
 }
 
+func (ni *NodeInfo) allocateIdleResource(ti *TaskInfo) error {
+	if ti.Resreq.LessEqual(ni.Idle) {
+		ni.Idle.Sub(ti.Resreq)
+		return nil
+	}
+	ni.State = NodeState{
+		Phase:  NotReady,
+		Reason: "OutOfSync",
+	}
+	return fmt.Errorf("Selected node NotReady")
+}
+
 // AddTask is used to add a task in nodeInfo object
 func (ni *NodeInfo) AddTask(task *TaskInfo) error {
 	key := PodKey(task.Pod)
@@ -173,12 +185,16 @@ func (ni *NodeInfo) AddTask(task *TaskInfo) error {
 	if ni.Node != nil {
 		switch ti.Status {
 		case Releasing:
+			if err := ni.allocateIdleResource(ti); err != nil {
+				return err
+			}
 			ni.Releasing.Add(ti.Resreq)
-			ni.Idle.Sub(ti.Resreq)
 		case Pipelined:
 			ni.Releasing.Sub(ti.Resreq)
 		default:
-			ni.Idle.Sub(ti.Resreq)
+			if err := ni.allocateIdleResource(ti); err != nil {
+				return err
+			}
 		}
 
 		ni.Used.Add(ti.Resreq)

pkg/scheduler/api/node_info_test.go

@@ -37,6 +37,9 @@ func TestNodeInfo_AddPod(t *testing.T) {
 	case01Node := buildNode("n1", buildResourceList("8000m", "10G"))
 	case01Pod1 := buildPod("c1", "p1", "n1", v1.PodRunning, buildResourceList("1000m", "1G"), []metav1.OwnerReference{}, make(map[string]string))
 	case01Pod2 := buildPod("c1", "p2", "n1", v1.PodRunning, buildResourceList("2000m", "2G"), []metav1.OwnerReference{}, make(map[string]string))
+	// case2
+	case02Node := buildNode("n2", buildResourceList("2000m", "1G"))
+	case02Pod1 := buildPod("c2", "p1", "n2", v1.PodUnknown, buildResourceList("1000m", "2G"), []metav1.OwnerReference{}, make(map[string]string))
 
 	tests := []struct {
 		name     string
@@ -63,6 +66,22 @@ func TestNodeInfo_AddPod(t *testing.T) {
 				},
 			},
 		},
+		{
+			name: "add 1 unknown pod",
+			node: case02Node,
+			pods: []*v1.Pod{case02Pod1},
+			expected: &NodeInfo{
+				Name:        "n2",
+				Node:        case02Node,
+				Idle:        buildResource("2000m", "1G"),
+				Used:        EmptyResource(),
+				Releasing:   EmptyResource(),
+				Allocatable: buildResource("2000m", "1G"),
+				Capability:  buildResource("2000m", "1G"),
+				State:       NodeState{Phase: NotReady, Reason: "OutOfSync"},
+				Tasks:       map[TaskID]*TaskInfo{},
+			},
+		},
 	}
 
 	for i, test := range tests {

test/kubemark/start-kubemark.sh

@@ -9,7 +9,7 @@ CRD_DIRECTORY="${VK_ROOT}/deployment/kube-batch/templates"
 QUEUE_DIR="${VK_ROOT}/config/queue"
 
 #Release version for kube batch
-RELEASE_VER=v0.4.2
+RELEASE_VER=v0.5
 
 #Ensure external cluster exists and kubectl binary works
 kubectl get nodes