You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
helm/pkg/kube/ready_test.go

537 lines
13 KiB

/*
Copyright The Helm Authors.
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
*/
package kube // import "helm.sh/helm/v3/pkg/kube"
import (
"context"
"testing"
appsv1 "k8s.io/api/apps/v1"
batchv1 "k8s.io/api/batch/v1"
corev1 "k8s.io/api/core/v1"
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
"k8s.io/apimachinery/pkg/runtime"
"k8s.io/apimachinery/pkg/util/intstr"
"k8s.io/client-go/kubernetes/fake"
)
const defaultNamespace = metav1.NamespaceDefault
func Test_ReadyChecker_deploymentReady(t *testing.T) {
type args struct {
rs *appsv1.ReplicaSet
dep *appsv1.Deployment
}
tests := []struct {
name string
args args
want bool
}{
{
name: "deployment is ready",
args: args{
rs: newReplicaSet("foo", 1, 1),
dep: newDeployment("foo", 1, 1, 0),
},
want: true,
},
{
name: "deployment is not ready",
args: args{
rs: newReplicaSet("foo", 0, 0),
dep: newDeployment("foo", 1, 1, 0),
},
want: false,
},
{
name: "deployment is ready when maxUnavailable is set",
args: args{
rs: newReplicaSet("foo", 2, 1),
dep: newDeployment("foo", 2, 1, 1),
},
want: true,
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
c := NewReadyChecker(fake.NewSimpleClientset(), nil)
if got := c.deploymentReady(tt.args.rs, tt.args.dep); got != tt.want {
t.Errorf("deploymentReady() = %v, want %v", got, tt.want)
}
})
}
}
func Test_ReadyChecker_daemonSetReady(t *testing.T) {
type args struct {
ds *appsv1.DaemonSet
}
tests := []struct {
name string
args args
want bool
}{
{
name: "daemonset is ready",
args: args{
ds: newDaemonSet("foo", 0, 1, 1, 1),
},
want: true,
},
{
name: "daemonset is not ready",
args: args{
ds: newDaemonSet("foo", 0, 0, 1, 1),
},
want: false,
},
{
name: "daemonset pods have not been scheduled successfully",
args: args{
ds: newDaemonSet("foo", 0, 0, 1, 0),
},
want: false,
},
{
name: "daemonset is ready when maxUnavailable is set",
args: args{
ds: newDaemonSet("foo", 1, 1, 2, 2),
},
want: true,
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
c := NewReadyChecker(fake.NewSimpleClientset(), nil)
if got := c.daemonSetReady(tt.args.ds); got != tt.want {
t.Errorf("daemonSetReady() = %v, want %v", got, tt.want)
}
})
}
}
func Test_ReadyChecker_statefulSetReady(t *testing.T) {
type args struct {
sts *appsv1.StatefulSet
}
tests := []struct {
name string
args args
want bool
}{
{
name: "statefulset is ready",
args: args{
sts: newStatefulSet("foo", 1, 0, 1, 1),
},
want: true,
},
{
name: "statefulset is not ready",
args: args{
sts: newStatefulSet("foo", 1, 0, 0, 1),
},
want: false,
},
{
name: "statefulset is ready when partition is specified",
args: args{
sts: newStatefulSet("foo", 2, 1, 2, 1),
},
want: true,
},
{
name: "statefulset is not ready when partition is set",
args: args{
sts: newStatefulSet("foo", 2, 1, 1, 0),
},
want: false,
},
{
name: "statefulset is ready when partition is set and no change in template",
args: args{
sts: newStatefulSet("foo", 2, 1, 2, 2),
},
want: true,
},
{
name: "statefulset is ready when partition is greater than replicas",
args: args{
sts: newStatefulSet("foo", 1, 2, 1, 1),
},
want: true,
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
c := NewReadyChecker(fake.NewSimpleClientset(), nil)
if got := c.statefulSetReady(tt.args.sts); got != tt.want {
t.Errorf("statefulSetReady() = %v, want %v", got, tt.want)
}
})
}
}
func Test_ReadyChecker_podsReadyForObject(t *testing.T) {
type args struct {
namespace string
obj runtime.Object
}
tests := []struct {
name string
args args
existPods []corev1.Pod
want bool
wantErr bool
}{
{
name: "pods ready for a replicaset",
args: args{
namespace: defaultNamespace,
obj: newReplicaSet("foo", 1, 1),
},
existPods: []corev1.Pod{
*newPodWithCondition("foo", corev1.ConditionTrue),
},
want: true,
wantErr: false,
},
{
name: "pods not ready for a replicaset",
args: args{
namespace: defaultNamespace,
obj: newReplicaSet("foo", 1, 1),
},
existPods: []corev1.Pod{
*newPodWithCondition("foo", corev1.ConditionFalse),
},
want: false,
wantErr: false,
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
c := NewReadyChecker(fake.NewSimpleClientset(), nil)
for _, pod := range tt.existPods {
if _, err := c.client.CoreV1().Pods(defaultNamespace).Create(context.TODO(), &pod, metav1.CreateOptions{}); err != nil {
t.Errorf("Failed to create Pod error: %v", err)
return
}
}
got, err := c.podsReadyForObject(context.TODO(), tt.args.namespace, tt.args.obj)
if (err != nil) != tt.wantErr {
t.Errorf("podsReadyForObject() error = %v, wantErr %v", err, tt.wantErr)
return
}
if got != tt.want {
t.Errorf("podsReadyForObject() got = %v, want %v", got, tt.want)
}
})
}
}
func Test_ReadyChecker_jobReady(t *testing.T) {
type args struct {
job *batchv1.Job
}
tests := []struct {
name string
args args
want bool
}{
{
name: "job is completed",
args: args{job: newJob("foo", 1, intToInt32(1), 1, 0)},
want: true,
},
{
name: "job is incomplete",
args: args{job: newJob("foo", 1, intToInt32(1), 0, 0)},
want: false,
},
{
name: "job is failed",
args: args{job: newJob("foo", 1, intToInt32(1), 0, 1)},
want: false,
},
{
name: "job is completed with retry",
args: args{job: newJob("foo", 1, intToInt32(1), 1, 1)},
want: true,
},
{
name: "job is failed with retry",
args: args{job: newJob("foo", 1, intToInt32(1), 0, 2)},
want: false,
},
{
name: "job is completed single run",
args: args{job: newJob("foo", 0, intToInt32(1), 1, 0)},
want: true,
},
{
name: "job is failed single run",
args: args{job: newJob("foo", 0, intToInt32(1), 0, 1)},
want: false,
},
{
name: "job with null completions",
args: args{job: newJob("foo", 0, nil, 1, 0)},
want: true,
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
c := NewReadyChecker(fake.NewSimpleClientset(), nil)
if got := c.jobReady(tt.args.job); got != tt.want {
t.Errorf("jobReady() = %v, want %v", got, tt.want)
}
})
}
}
func Test_ReadyChecker_volumeReady(t *testing.T) {
type args struct {
v *corev1.PersistentVolumeClaim
}
tests := []struct {
name string
args args
want bool
}{
{
name: "pvc is bound",
args: args{
v: newPersistentVolumeClaim("foo", corev1.ClaimBound),
},
want: true,
},
{
name: "pvc is not ready",
args: args{
v: newPersistentVolumeClaim("foo", corev1.ClaimPending),
},
want: false,
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
c := NewReadyChecker(fake.NewSimpleClientset(), nil)
if got := c.volumeReady(tt.args.v); got != tt.want {
t.Errorf("volumeReady() = %v, want %v", got, tt.want)
}
})
}
}
func newDaemonSet(name string, maxUnavailable, numberReady, desiredNumberScheduled, updatedNumberScheduled int) *appsv1.DaemonSet {
return &appsv1.DaemonSet{
ObjectMeta: metav1.ObjectMeta{
Name: name,
Namespace: defaultNamespace,
},
Spec: appsv1.DaemonSetSpec{
UpdateStrategy: appsv1.DaemonSetUpdateStrategy{
Type: appsv1.RollingUpdateDaemonSetStrategyType,
RollingUpdate: &appsv1.RollingUpdateDaemonSet{
MaxUnavailable: func() *intstr.IntOrString { i := intstr.FromInt(maxUnavailable); return &i }(),
},
},
Selector: &metav1.LabelSelector{MatchLabels: map[string]string{"name": name}},
Template: corev1.PodTemplateSpec{
ObjectMeta: metav1.ObjectMeta{
Name: name,
Labels: map[string]string{"name": name},
},
Spec: corev1.PodSpec{
Containers: []corev1.Container{
{
Image: "nginx",
},
},
},
},
},
Status: appsv1.DaemonSetStatus{
DesiredNumberScheduled: int32(desiredNumberScheduled),
NumberReady: int32(numberReady),
UpdatedNumberScheduled: int32(updatedNumberScheduled),
},
}
}
func newStatefulSet(name string, replicas, partition, readyReplicas, updatedReplicas int) *appsv1.StatefulSet {
return &appsv1.StatefulSet{
ObjectMeta: metav1.ObjectMeta{
Name: name,
Namespace: defaultNamespace,
},
Spec: appsv1.StatefulSetSpec{
UpdateStrategy: appsv1.StatefulSetUpdateStrategy{
Type: appsv1.RollingUpdateStatefulSetStrategyType,
RollingUpdate: &appsv1.RollingUpdateStatefulSetStrategy{
Partition: intToInt32(partition),
},
},
Replicas: intToInt32(replicas),
Selector: &metav1.LabelSelector{MatchLabels: map[string]string{"name": name}},
Template: corev1.PodTemplateSpec{
ObjectMeta: metav1.ObjectMeta{
Name: name,
Labels: map[string]string{"name": name},
},
Spec: corev1.PodSpec{
Containers: []corev1.Container{
{
Image: "nginx",
},
},
},
},
},
Status: appsv1.StatefulSetStatus{
UpdatedReplicas: int32(updatedReplicas),
ReadyReplicas: int32(readyReplicas),
},
}
}
func newDeployment(name string, replicas, maxSurge, maxUnavailable int) *appsv1.Deployment {
return &appsv1.Deployment{
ObjectMeta: metav1.ObjectMeta{
Name: name,
Namespace: defaultNamespace,
},
Spec: appsv1.DeploymentSpec{
Strategy: appsv1.DeploymentStrategy{
Type: appsv1.RollingUpdateDeploymentStrategyType,
RollingUpdate: &appsv1.RollingUpdateDeployment{
MaxUnavailable: func() *intstr.IntOrString { i := intstr.FromInt(maxUnavailable); return &i }(),
MaxSurge: func() *intstr.IntOrString { i := intstr.FromInt(maxSurge); return &i }(),
},
},
Replicas: intToInt32(replicas),
Selector: &metav1.LabelSelector{MatchLabels: map[string]string{"name": name}},
Template: corev1.PodTemplateSpec{
ObjectMeta: metav1.ObjectMeta{
Name: name,
Labels: map[string]string{"name": name},
},
Spec: corev1.PodSpec{
Containers: []corev1.Container{
{
Image: "nginx",
},
},
},
},
},
}
}
func newReplicaSet(name string, replicas int, readyReplicas int) *appsv1.ReplicaSet {
d := newDeployment(name, replicas, 0, 0)
return &appsv1.ReplicaSet{
ObjectMeta: metav1.ObjectMeta{
Name: name,
Namespace: defaultNamespace,
Labels: d.Spec.Selector.MatchLabels,
OwnerReferences: []metav1.OwnerReference{*metav1.NewControllerRef(d, d.GroupVersionKind())},
},
Spec: appsv1.ReplicaSetSpec{
Selector: d.Spec.Selector,
Replicas: intToInt32(replicas),
Template: d.Spec.Template,
},
Status: appsv1.ReplicaSetStatus{
ReadyReplicas: int32(readyReplicas),
},
}
}
func newPodWithCondition(name string, podReadyCondition corev1.ConditionStatus) *corev1.Pod {
return &corev1.Pod{
ObjectMeta: metav1.ObjectMeta{
Name: name,
Namespace: defaultNamespace,
Labels: map[string]string{"name": name},
},
Spec: corev1.PodSpec{
Containers: []corev1.Container{
{
Image: "nginx",
},
},
},
Status: corev1.PodStatus{
Conditions: []corev1.PodCondition{
{
Type: corev1.PodReady,
Status: podReadyCondition,
},
},
},
}
}
func newPersistentVolumeClaim(name string, phase corev1.PersistentVolumeClaimPhase) *corev1.PersistentVolumeClaim {
return &corev1.PersistentVolumeClaim{
ObjectMeta: metav1.ObjectMeta{
Name: name,
Namespace: defaultNamespace,
},
Status: corev1.PersistentVolumeClaimStatus{
Phase: phase,
},
}
}
func newJob(name string, backoffLimit int, completions *int32, succeeded int, failed int) *batchv1.Job {
return &batchv1.Job{
ObjectMeta: metav1.ObjectMeta{
Name: name,
Namespace: defaultNamespace,
},
Spec: batchv1.JobSpec{
BackoffLimit: intToInt32(backoffLimit),
Completions: completions,
Template: corev1.PodTemplateSpec{
ObjectMeta: metav1.ObjectMeta{
Name: name,
Labels: map[string]string{"name": name},
},
Spec: corev1.PodSpec{
Containers: []corev1.Container{
{
Image: "nginx",
},
},
},
},
},
Status: batchv1.JobStatus{
Succeeded: int32(succeeded),
Failed: int32(failed),
},
}
}
func intToInt32(i int) *int32 {
i32 := int32(i)
return &i32
}