Skip to content

Add E2E conformance test for BroadcastJob uncordon handling #2090

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.

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 5 commits into
base: master
Choose a base branch
from
Open
Changes from all commits
Commits
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
90 changes: 85 additions & 5 deletions test/e2e/apps/broadcastjob.go
Original file line number Diff line number Diff line change
Expand Up @@ -20,20 +20,19 @@ import (
"context"
"time"

"k8s.io/client-go/util/retry"

"github.com/onsi/ginkgo"
"github.com/onsi/gomega"
appsv1alpha1 "github.com/openkruise/kruise/apis/apps/v1alpha1"
kruiseclientset "github.com/openkruise/kruise/pkg/client/clientset/versioned"
"github.com/openkruise/kruise/pkg/util"
"github.com/openkruise/kruise/test/e2e/framework"

"github.com/onsi/ginkgo"
"github.com/onsi/gomega"
v1 "k8s.io/api/core/v1"
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
"k8s.io/apimachinery/pkg/types"
"k8s.io/apimachinery/pkg/util/intstr"
"k8s.io/apimachinery/pkg/util/rand"
clientset "k8s.io/client-go/kubernetes"
"k8s.io/client-go/util/retry"
"k8s.io/utils/integer"
)

Expand Down Expand Up @@ -156,4 +155,85 @@ var _ = SIGDescribe("BroadcastJob", func() {
}, 60*time.Second, time.Second).Should(gomega.Equal(int32(len(nodes))))
})
})

framework.KruiseDescribe("BroadcastJob uncordon handling", func() {
framework.ConformanceIt("creates missing pod after node uncordon", func() {
// Create fake node
fakeNode, err := nodeTester.CreateFakeNode(randStr)
gomega.Expect(err).NotTo(gomega.HaveOccurred())

// Cordon the fake node
ginkgo.By("Cordoning fake node " + fakeNode.Name)
_, err = c.CoreV1().Nodes().Patch(context.TODO(), fakeNode.Name,
types.StrategicMergePatchType,
[]byte(`{"spec":{"unschedulable":true}}`),
metav1.PatchOptions{})
gomega.Expect(err).NotTo(gomega.HaveOccurred())

// Create BroadcastJob job-" + randStr
job := &appsv1alpha1.BroadcastJob{
ObjectMeta: metav1.ObjectMeta{Namespace: ns, Name: "job-" + randStr},
Spec: appsv1alpha1.BroadcastJobSpec{
Template: v1.PodTemplateSpec{
Spec: v1.PodSpec{
Tolerations: []v1.Toleration{{Key: framework.E2eFakeKey, Operator: v1.TolerationOpEqual, Value: randStr, Effect: v1.TaintEffectNoSchedule}},
Containers: []v1.Container{{
Name: "box",
Image: BusyboxImage,
Command: []string{"/bin/sh", "-c", "sleep 30"},
}},
RestartPolicy: v1.RestartPolicyNever,
},
},
CompletionPolicy: appsv1alpha1.CompletionPolicy{Type: appsv1alpha1.Always},
},
}

nodes, err := nodeTester.ListRealNodesWithFake(job.Spec.Template.Spec.Tolerations)
gomega.Expect(err).NotTo(gomega.HaveOccurred())
totalNodes := int32(len(nodes))
parallelism := intstr.FromInt(len(nodes))
job.Spec.Parallelism = &parallelism

ginkgo.By("Creating BroadcastJob " + job.Name)
job, err = tester.CreateBroadcastJob(job)
gomega.Expect(err).NotTo(gomega.HaveOccurred())

ginkgo.By("Verify desired count equals total nodes - 1 (due to cordoned node)")
gomega.Eventually(func() int32 {
job, err = tester.GetBroadcastJob(job.Name)
gomega.Expect(err).NotTo(gomega.HaveOccurred())
return job.Status.Desired
}, 30*time.Second, time.Second).Should(gomega.Equal(totalNodes - 1))

ginkgo.By("Verify active pods equals total nodes - 1 (as pod is not created on cordoned node)")
gomega.Eventually(func() int32 {
job, err = tester.GetBroadcastJob(job.Name)
gomega.Expect(err).NotTo(gomega.HaveOccurred())
return job.Status.Active
}, 60*time.Second, 3*time.Second).Should(gomega.Equal(totalNodes - 1))

// Uncordon the fake node
ginkgo.By("Uncordoning fake node " + fakeNode.Name)
_, err = c.CoreV1().Nodes().Patch(context.TODO(), fakeNode.Name,
types.StrategicMergePatchType,
[]byte(`{"spec":{"unschedulable":false}}`),
metav1.PatchOptions{})
gomega.Expect(err).NotTo(gomega.HaveOccurred())

ginkgo.By("Verify desired count becomes total nodes after uncordon")
gomega.Eventually(func() int32 {
job, err = tester.GetBroadcastJob(job.Name)
gomega.Expect(err).NotTo(gomega.HaveOccurred())
return job.Status.Desired
}, 30*time.Second, time.Second).Should(gomega.Equal(totalNodes))

ginkgo.By("Verify active pods becomes total nodes after uncordon (missing pod now created)")
gomega.Eventually(func() int32 {
job, err = tester.GetBroadcastJob(job.Name)
gomega.Expect(err).NotTo(gomega.HaveOccurred())
return job.Status.Active
}, 60*time.Second, 3*time.Second).Should(gomega.Equal(totalNodes))
})
})
})
Loading