Skip to content

fix: support rolling update lws with 1 replica #491

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
+148 −29
Open

fix: support rolling update lws with 1 replica #491

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
13 changes: 7 additions & 6 deletions pkg/controllers/leaderworkerset_controller.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -271,11 +271,12 @@ func (r *LeaderWorkerSetReconciler) rollingUpdateParameters(ctx context.Context,

// wantReplicas calculates the final replicas if needed.
wantReplicas := func(unreadyReplicas int32) int32 {
if unreadyReplicas <= int32(maxSurge) {
// When we have n unready replicas and n bursted replicas, we should
// start to release the burst replica gradually for the accommodation of
// the unready ones.
finalReplicas := lwsReplicas + utils.NonZeroValue(int32(unreadyReplicas)-1)
if unreadyReplicas < int32(maxSurge) {
// When the unready replicas less than maxSurge, we should start to release
// the burst replica gradually for the accommodation of the unready ones.
// Actually we should keep the burst replicas when the unready replicas is
// equal to maxSurge, because the rolling update is not completed yet.
finalReplicas := lwsReplicas + utils.NonZeroValue(int32(unreadyReplicas))
r.Record.Eventf(lws, corev1.EventTypeNormal, GroupsProgressing, fmt.Sprintf("deleting surge replica %s-%d", lws.Name, finalReplicas))
return finalReplicas
}
Expand All @@ -286,7 +287,7 @@ func (r *LeaderWorkerSetReconciler) rollingUpdateParameters(ctx context.Context,
// Indicates a new rolling update here.
if leaderWorkerSetUpdated {
// Processing scaling up/down first prior to rolling update.
return min(lwsReplicas, stsReplicas), wantReplicas(lwsReplicas), nil
return min(lwsReplicas, stsReplicas), lwsReplicas, nil
}

partition := *sts.Spec.UpdateStrategy.RollingUpdate.Partition
Expand Down
4 changes: 1 addition & 3 deletions pkg/controllers/leaderworkerset_controller_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -29,10 +29,8 @@ import (
coreapplyv1 "k8s.io/client-go/applyconfigurations/core/v1"
metaapplyv1 "k8s.io/client-go/applyconfigurations/meta/v1"
"k8s.io/utils/ptr"

leaderworkerset "sigs.k8s.io/lws/api/leaderworkerset/v1"

"sigs.k8s.io/controller-runtime/pkg/client/fake"
leaderworkerset "sigs.k8s.io/lws/api/leaderworkerset/v1"
revisionutils "sigs.k8s.io/lws/pkg/utils/revision"
"sigs.k8s.io/lws/test/wrappers"
)
Expand Down
8 changes: 4 additions & 4 deletions test/e2e/e2e_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -140,7 +140,7 @@ var _ = ginkgo.Describe("leaderWorkerSet e2e tests", func() {
testing.UpdateWorkerTemplate(ctx, k8sClient, lws)

// Happen during rolling update.
testing.ExpectValidLeaderStatefulSet(ctx, k8sClient, lws, 7)
testing.ExpectValidLeaderStatefulSet(ctx, k8sClient, lws, 8)

// Rolling update completes.
testing.ExpectValidLeaderStatefulSet(ctx, k8sClient, lws, 4)
Expand Down Expand Up @@ -186,7 +186,7 @@ var _ = ginkgo.Describe("leaderWorkerSet e2e tests", func() {
testing.UpdateWorkerTemplate(ctx, k8sClient, lws)

// Happen during rolling update.
testing.ExpectValidLeaderStatefulSet(ctx, k8sClient, lws, 7)
testing.ExpectValidLeaderStatefulSet(ctx, k8sClient, lws, 8)

// Rolling update completes.
testing.ExpectValidLeaderStatefulSet(ctx, k8sClient, lws, 4)
Expand Down Expand Up @@ -245,8 +245,8 @@ var _ = ginkgo.Describe("leaderWorkerSet e2e tests", func() {

testing.UpdateWorkerTemplate(ctx, k8sClient, lws)

testing.ExpectValidLeaderStatefulSet(ctx, k8sClient, lws, 7)
testing.ExpectValidServices(ctx, k8sClient, lws, 7)
testing.ExpectValidLeaderStatefulSet(ctx, k8sClient, lws, 8)
testing.ExpectValidServices(ctx, k8sClient, lws, 8)
// Rolling update completes.
testing.ExpectValidLeaderStatefulSet(ctx, k8sClient, lws, 4)
testing.ExpectValidWorkerStatefulSets(ctx, lws, k8sClient, true)
Expand Down
112 changes: 96 additions & 16 deletions test/integration/controllers/leaderworkerset_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1061,22 +1061,22 @@ var _ = ginkgo.Describe("LeaderWorkerSet controller", func() {
// Rolling update index-1 replica.
lwsUpdateFn: func(lws *leaderworkerset.LeaderWorkerSet) {
testing.SetPodGroupToReady(ctx, k8sClient, lws.Name+"-1", lws)
// Reclaim the replica.
testing.DeleteLeaderPod(ctx, k8sClient, lws, 4, 5)
},
checkLWSState: func(lws *leaderworkerset.LeaderWorkerSet) {
testing.ExpectValidLeaderStatefulSet(ctx, k8sClient, lws, 4)
testing.ExpectValidLeaderStatefulSet(ctx, k8sClient, lws, 5)
testing.ExpectLeaderWorkerSetUnavailable(ctx, k8sClient, lws, "All replicas are ready")
testing.ExpectLeaderWorkerSetProgressing(ctx, k8sClient, lws, "Replicas are progressing")
testing.ExpectLeaderWorkerSetUpgradeInProgress(ctx, k8sClient, lws, "Rolling Upgrade is in progress")
testing.ExpectStatefulsetPartitionEqualTo(ctx, k8sClient, lws, 0)
testing.ExpectLeaderWorkerSetStatusReplicas(ctx, k8sClient, lws, 4, 3)
testing.ExpectLeaderWorkerSetStatusReplicas(ctx, k8sClient, lws, 5, 4)
},
},
{
// Rolling update index-0 replica.
lwsUpdateFn: func(lws *leaderworkerset.LeaderWorkerSet) {
testing.SetPodGroupToReady(ctx, k8sClient, lws.Name+"-0", lws)
// Reclaim the replica.
testing.DeleteLeaderPod(ctx, k8sClient, lws, 4, 5)
},
checkLWSState: func(lws *leaderworkerset.LeaderWorkerSet) {
testing.ExpectValidLeaderStatefulSet(ctx, k8sClient, lws, 4)
Expand Down Expand Up @@ -1159,6 +1159,83 @@ var _ = ginkgo.Describe("LeaderWorkerSet controller", func() {
},
},
}),
ginkgo.Entry("rolling update with one replicas", &testCase{
makeLeaderWorkerSet: func(nsName string) *wrappers.LeaderWorkerSetWrapper {
return wrappers.BuildLeaderWorkerSet(nsName).MaxSurge(1).Replica(1)
},
updates: []*update{
{
// Set lws to available condition.
lwsUpdateFn: func(lws *leaderworkerset.LeaderWorkerSet) {
testing.SetPodGroupsToReady(ctx, k8sClient, lws, 1)
},
checkLWSState: func(lws *leaderworkerset.LeaderWorkerSet) {
testing.ExpectValidLeaderStatefulSet(ctx, k8sClient, lws, 1)
testing.ExpectValidWorkerStatefulSets(ctx, lws, k8sClient, true)
testing.ExpectLeaderWorkerSetAvailable(ctx, k8sClient, lws, "All replicas are ready")
testing.ExpectStatefulsetPartitionEqualTo(ctx, k8sClient, lws, 0)
testing.ExpectLeaderWorkerSetStatusReplicas(ctx, k8sClient, lws, 1, 1)
},
},
{
// Update the worker template.
lwsUpdateFn: func(lws *leaderworkerset.LeaderWorkerSet) {
gomega.Eventually(func() error {
var leaderworkerset leaderworkerset.LeaderWorkerSet
if err := k8sClient.Get(ctx, types.NamespacedName{Name: lws.Name, Namespace: lws.Namespace}, &leaderworkerset); err != nil {
return err
}
leaderworkerset.Spec.LeaderWorkerTemplate.WorkerTemplate.Spec.Containers[0].Name = "new-worker"
return k8sClient.Update(ctx, &leaderworkerset)
}, testing.Timeout, testing.Interval).Should(gomega.Succeed())

var leaderSts appsv1.StatefulSet
testing.GetLeaderStatefulset(ctx, lws, k8sClient, &leaderSts)
// Create leader pod for maxSurge.
gomega.Expect(testing.CreateLeaderPods(ctx, leaderSts, k8sClient, lws, 1, 2)).To(gomega.Succeed())
},
checkLWSState: func(lws *leaderworkerset.LeaderWorkerSet) {
testing.ExpectValidLeaderStatefulSet(ctx, k8sClient, lws, 2)
testing.ExpectLeaderWorkerSetUnavailable(ctx, k8sClient, lws, "All replicas are ready")
testing.ExpectLeaderWorkerSetProgressing(ctx, k8sClient, lws, "Replicas are progressing")
testing.ExpectLeaderWorkerSetUpgradeInProgress(ctx, k8sClient, lws, "Rolling Upgrade is in progress")
testing.ExpectStatefulsetPartitionEqualTo(ctx, k8sClient, lws, 0)
},
},
{
// rolling update the last replica
lwsUpdateFn: func(lws *leaderworkerset.LeaderWorkerSet) {
testing.SetPodGroupToReady(ctx, k8sClient, lws.Name+"-1", lws)
},
checkLWSState: func(lws *leaderworkerset.LeaderWorkerSet) {
testing.ExpectValidLeaderStatefulSet(ctx, k8sClient, lws, 2)
testing.ExpectLeaderWorkerSetUnavailable(ctx, k8sClient, lws, "All replicas are ready")
testing.ExpectLeaderWorkerSetProgressing(ctx, k8sClient, lws, "Replicas are progressing")
testing.ExpectLeaderWorkerSetUpgradeInProgress(ctx, k8sClient, lws, "Rolling Upgrade is in progress")
testing.ExpectStatefulsetPartitionEqualTo(ctx, k8sClient, lws, 0)
testing.ExpectLeaderWorkerSetStatusReplicas(ctx, k8sClient, lws, 2, 1)
},
},

{
// rolling update the first replica
lwsUpdateFn: func(lws *leaderworkerset.LeaderWorkerSet) {
testing.SetPodGroupToReady(ctx, k8sClient, lws.Name+"-0", lws)
// Reclaim the replica.
testing.DeleteLeaderPod(ctx, k8sClient, lws, 1, 2)
},
checkLWSState: func(lws *leaderworkerset.LeaderWorkerSet) {
testing.ExpectValidLeaderStatefulSet(ctx, k8sClient, lws, 1)
testing.ExpectValidWorkerStatefulSets(ctx, lws, k8sClient, true)
testing.ExpectLeaderWorkerSetAvailable(ctx, k8sClient, lws, "All replicas are ready")
testing.ExpectLeaderWorkerSetNotProgressing(ctx, k8sClient, lws, "Replicas are progressing")
testing.ExpectLeaderWorkerSetNoUpgradeInProgress(ctx, k8sClient, lws, "Rolling Upgrade is in progress")
testing.ExpectStatefulsetPartitionEqualTo(ctx, k8sClient, lws, 0)
testing.ExpectLeaderWorkerSetStatusReplicas(ctx, k8sClient, lws, 1, 1)
},
},
},
}),
ginkgo.Entry("rolling update with replicas scaled down and maxSurge set", &testCase{
makeLeaderWorkerSet: func(nsName string) *wrappers.LeaderWorkerSetWrapper {
return wrappers.BuildLeaderWorkerSet(nsName).Replica(6).MaxSurge(2)
Expand Down Expand Up @@ -1258,10 +1335,10 @@ var _ = ginkgo.Describe("LeaderWorkerSet controller", func() {
var leaderSts appsv1.StatefulSet
testing.GetLeaderStatefulset(ctx, lws, k8sClient, &leaderSts)
// Create leader pod for maxSurge.
gomega.Expect(testing.CreateLeaderPods(ctx, leaderSts, k8sClient, lws, 2, 3)).To(gomega.Succeed())
gomega.Expect(testing.CreateLeaderPods(ctx, leaderSts, k8sClient, lws, 2, 4)).To(gomega.Succeed())
},
checkLWSState: func(lws *leaderworkerset.LeaderWorkerSet) {
testing.ExpectValidLeaderStatefulSet(ctx, k8sClient, lws, 3)
testing.ExpectValidLeaderStatefulSet(ctx, k8sClient, lws, 4)
testing.ExpectLeaderWorkerSetUnavailable(ctx, k8sClient, lws, "All replicas are ready")
testing.ExpectLeaderWorkerSetProgressing(ctx, k8sClient, lws, "Replicas are progressing")
testing.ExpectLeaderWorkerSetUpgradeInProgress(ctx, k8sClient, lws, "Rolling Upgrade is in progress")
Expand All @@ -1271,7 +1348,7 @@ var _ = ginkgo.Describe("LeaderWorkerSet controller", func() {
{
// Set all groups to ready.
lwsUpdateFn: func(lws *leaderworkerset.LeaderWorkerSet) {
testing.SetPodGroupsToReady(ctx, k8sClient, lws, 3)
testing.SetPodGroupsToReady(ctx, k8sClient, lws, 4)
},
checkLWSState: func(lws *leaderworkerset.LeaderWorkerSet) {
testing.ExpectLeaderWorkerSetAvailable(ctx, k8sClient, lws, "All replicas are ready")
Expand Down Expand Up @@ -1393,52 +1470,55 @@ var _ = ginkgo.Describe("LeaderWorkerSet controller", func() {
return k8sClient.Update(ctx, &leaderworkerset)
}, testing.Timeout, testing.Interval).Should(gomega.Succeed())
testing.DeleteLeaderPod(ctx, k8sClient, lws, 4, 8)
// Reclaim the last replica.
testing.DeleteLeaderPod(ctx, k8sClient, lws, 3, 4)
},
checkLWSState: func(lws *leaderworkerset.LeaderWorkerSet) {
testing.ExpectValidLeaderStatefulSet(ctx, k8sClient, lws, 3)
testing.ExpectValidLeaderStatefulSet(ctx, k8sClient, lws, 4)
testing.ExpectLeaderWorkerSetUnavailable(ctx, k8sClient, lws, "All replicas are ready")
testing.ExpectLeaderWorkerSetProgressing(ctx, k8sClient, lws, "Replicas are progressing")
testing.ExpectLeaderWorkerSetUpgradeInProgress(ctx, k8sClient, lws, "Rolling Upgrade is in progress")
testing.ExpectStatefulsetPartitionEqualTo(ctx, k8sClient, lws, 1)
testing.ExpectLeaderWorkerSetStatusReplicas(ctx, k8sClient, lws, 3, 0)
testing.ExpectLeaderWorkerSetStatusReplicas(ctx, k8sClient, lws, 4, 0)
},
},
{
// Rolling update index-2 replica.
lwsUpdateFn: func(lws *leaderworkerset.LeaderWorkerSet) {
testing.SetPodGroupToReady(ctx, k8sClient, lws.Name+"-2", lws)
testing.SetPodGroupToReady(ctx, k8sClient, lws.Name+"-3", lws)
// Reclaim the last replica.
// testing.DeleteLeaderPod(ctx, k8sClient, lws, 3, 4)
},
checkLWSState: func(lws *leaderworkerset.LeaderWorkerSet) {
testing.ExpectValidLeaderStatefulSet(ctx, k8sClient, lws, 3)
testing.ExpectValidLeaderStatefulSet(ctx, k8sClient, lws, 4)
testing.ExpectLeaderWorkerSetUnavailable(ctx, k8sClient, lws, "All replicas are ready")
testing.ExpectLeaderWorkerSetProgressing(ctx, k8sClient, lws, "Replicas are progressing")
testing.ExpectLeaderWorkerSetUpgradeInProgress(ctx, k8sClient, lws, "Rolling Upgrade is in progress")
testing.ExpectStatefulsetPartitionEqualTo(ctx, k8sClient, lws, 0)
testing.ExpectLeaderWorkerSetStatusReplicas(ctx, k8sClient, lws, 3, 1)
testing.ExpectLeaderWorkerSetStatusReplicas(ctx, k8sClient, lws, 4, 2)
},
},
{
// Rolling update index-1 replica.
lwsUpdateFn: func(lws *leaderworkerset.LeaderWorkerSet) {
testing.SetPodGroupToReady(ctx, k8sClient, lws.Name+"-1", lws)
// Reclaim the last replica.
testing.DeleteLeaderPod(ctx, k8sClient, lws, 2, 3)
testing.DeleteLeaderPod(ctx, k8sClient, lws, 3, 4)
},
checkLWSState: func(lws *leaderworkerset.LeaderWorkerSet) {
testing.ExpectValidLeaderStatefulSet(ctx, k8sClient, lws, 2)
testing.ExpectValidLeaderStatefulSet(ctx, k8sClient, lws, 3)
testing.ExpectLeaderWorkerSetUnavailable(ctx, k8sClient, lws, "All replicas are ready")
testing.ExpectLeaderWorkerSetProgressing(ctx, k8sClient, lws, "Replicas are progressing")
testing.ExpectLeaderWorkerSetUpgradeInProgress(ctx, k8sClient, lws, "Rolling Upgrade is in progress")
testing.ExpectStatefulsetPartitionEqualTo(ctx, k8sClient, lws, 0)
testing.ExpectLeaderWorkerSetStatusReplicas(ctx, k8sClient, lws, 2, 1)
testing.ExpectLeaderWorkerSetStatusReplicas(ctx, k8sClient, lws, 3, 2)
},
},
{
// Rolling update index-0 replica.
lwsUpdateFn: func(lws *leaderworkerset.LeaderWorkerSet) {
testing.SetPodGroupToReady(ctx, k8sClient, lws.Name+"-0", lws)
// Reclaim the last replica.
testing.DeleteLeaderPod(ctx, k8sClient, lws, 2, 3)
},
checkLWSState: func(lws *leaderworkerset.LeaderWorkerSet) {
testing.ExpectValidLeaderStatefulSet(ctx, k8sClient, lws, 2)
Expand Down
40 changes: 40 additions & 0 deletions test/wrappers/wrappers.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -18,6 +18,7 @@ import (
"fmt"
"strconv"

appsv1 "k8s.io/api/apps/v1"
corev1 "k8s.io/api/core/v1"
"k8s.io/apimachinery/pkg/api/resource"
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
Expand Down Expand Up @@ -197,6 +198,45 @@ func MakePodWithLabels(setName, groupIndex, workerIndex, namespace string, size
}
}

func MakePodWithLabelsAndStatus(setName, groupIndex, workerIndex, namespace string, size int, status corev1.PodPhase) *corev1.Pod {
pod := MakePodWithLabels(setName, groupIndex, workerIndex, namespace, size)
pod.Status.Phase = status

if status == corev1.PodRunning {
pod.Status.Conditions = []corev1.PodCondition{
{
Type: corev1.PodReady,
Status: corev1.ConditionTrue,
},
}
}
return pod
}

func MakeLeaderStatefulSetWithLabels(setName, namespace string, replica int) *appsv1.StatefulSet {
return &appsv1.StatefulSet{
ObjectMeta: metav1.ObjectMeta{
Name: setName,
Namespace: namespace,
Labels: map[string]string{
leaderworkerset.SetNameLabelKey: setName,
},
Annotations: map[string]string{
leaderworkerset.ReplicasAnnotationKey: strconv.Itoa(replica),
},
},
Spec: appsv1.StatefulSetSpec{
Replicas: ptr.To(int32(replica)),
UpdateStrategy: appsv1.StatefulSetUpdateStrategy{
Type: appsv1.RollingUpdateStatefulSetStrategyType,
RollingUpdate: &appsv1.RollingUpdateStatefulSetStrategy{
Partition: ptr.To[int32](0),
},
},
},
}
}

func MakeWorkerPodSpec() corev1.PodSpec {
return corev1.PodSpec{
Containers: []corev1.Container{
Expand Down