代码拉取完成,页面将自动刷新
/*
Copyright 2018 The Kubernetes 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 storage
import (
"time"
. "github.com/onsi/ginkgo"
. "github.com/onsi/gomega"
apps "k8s.io/api/apps/v1"
"k8s.io/api/core/v1"
storage "k8s.io/api/storage/v1"
"k8s.io/apimachinery/pkg/api/resource"
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
utilerrors "k8s.io/apimachinery/pkg/util/errors"
"k8s.io/apimachinery/pkg/util/wait"
clientset "k8s.io/client-go/kubernetes"
"k8s.io/kubernetes/pkg/client/conditions"
"k8s.io/kubernetes/test/e2e/framework"
"k8s.io/kubernetes/test/e2e/storage/utils"
)
var _ = utils.SIGDescribe("Mounted volume expand[Slow]", func() {
var (
c clientset.Interface
ns string
err error
pvc *v1.PersistentVolumeClaim
resizableSc *storage.StorageClass
nodeName string
isNodeLabeled bool
nodeKeyValueLabel map[string]string
nodeLabelValue string
nodeKey string
)
f := framework.NewDefaultFramework("mounted-volume-expand")
BeforeEach(func() {
framework.SkipUnlessProviderIs("aws", "gce")
c = f.ClientSet
ns = f.Namespace.Name
framework.ExpectNoError(framework.WaitForAllNodesSchedulable(c, framework.TestContext.NodeSchedulableTimeout))
nodeList := framework.GetReadySchedulableNodesOrDie(f.ClientSet)
if len(nodeList.Items) != 0 {
nodeName = nodeList.Items[0].Name
} else {
framework.Failf("Unable to find ready and schedulable Node")
}
nodeKey = "mounted_volume_expand"
if !isNodeLabeled {
nodeLabelValue = ns
nodeKeyValueLabel = make(map[string]string)
nodeKeyValueLabel[nodeKey] = nodeLabelValue
framework.AddOrUpdateLabelOnNode(c, nodeName, nodeKey, nodeLabelValue)
isNodeLabeled = true
}
test := storageClassTest{
name: "default",
claimSize: "2Gi",
}
resizableSc, err = createResizableStorageClass(test, ns, "resizing", c)
Expect(err).NotTo(HaveOccurred(), "Error creating resizable storage class")
Expect(*resizableSc.AllowVolumeExpansion).To(BeTrue())
pvc = newClaim(test, ns, "default")
pvc.Spec.StorageClassName = &resizableSc.Name
pvc, err = c.CoreV1().PersistentVolumeClaims(pvc.Namespace).Create(pvc)
Expect(err).NotTo(HaveOccurred(), "Error creating pvc")
})
framework.AddCleanupAction(func() {
if len(nodeLabelValue) > 0 {
framework.RemoveLabelOffNode(c, nodeName, nodeKey)
}
})
AfterEach(func() {
framework.Logf("AfterEach: Cleaning up resources for mounted volume resize")
if c != nil {
if errs := framework.PVPVCCleanup(c, ns, nil, pvc); len(errs) > 0 {
framework.Failf("AfterEach: Failed to delete PVC and/or PV. Errors: %v", utilerrors.NewAggregate(errs))
}
pvc, nodeName, isNodeLabeled, nodeLabelValue = nil, "", false, ""
nodeKeyValueLabel = make(map[string]string)
}
})
It("Should verify mounted devices can be resized", func() {
By("Waiting for PVC to be in bound phase")
pvcClaims := []*v1.PersistentVolumeClaim{pvc}
pvs, err := framework.WaitForPVClaimBoundPhase(c, pvcClaims, framework.ClaimProvisionTimeout)
Expect(err).NotTo(HaveOccurred(), "Failed waiting for PVC to be bound %v", err)
Expect(len(pvs)).To(Equal(1))
By("Creating a deployment with the provisioned volume")
deployment, err := framework.CreateDeployment(c, int32(1), map[string]string{"test": "app"}, nodeKeyValueLabel, ns, pvcClaims, "")
Expect(err).NotTo(HaveOccurred(), "Failed creating deployment %v", err)
defer c.AppsV1().Deployments(ns).Delete(deployment.Name, &metav1.DeleteOptions{})
By("Expanding current pvc")
newSize := resource.MustParse("6Gi")
pvc, err = expandPVCSize(pvc, newSize, c)
Expect(err).NotTo(HaveOccurred(), "While updating pvc for more size")
Expect(pvc).NotTo(BeNil())
pvcSize := pvc.Spec.Resources.Requests[v1.ResourceStorage]
if pvcSize.Cmp(newSize) != 0 {
framework.Failf("error updating pvc size %q", pvc.Name)
}
By("Waiting for cloudprovider resize to finish")
err = waitForControllerVolumeResize(pvc, c)
Expect(err).NotTo(HaveOccurred(), "While waiting for pvc resize to finish")
By("Getting a pod from deployment")
podList, err := framework.GetPodsForDeployment(c, deployment)
Expect(podList.Items).NotTo(BeEmpty())
pod := podList.Items[0]
By("Deleting the pod from deployment")
err = framework.DeletePodWithWait(f, c, &pod)
Expect(err).NotTo(HaveOccurred(), "while deleting pod for resizing")
By("Waiting for deployment to create new pod")
pod, err = waitForDeploymentToRecreatePod(c, deployment)
Expect(err).NotTo(HaveOccurred(), "While waiting for pod to be recreated")
By("Waiting for file system resize to finish")
pvc, err = waitForFSResize(pvc, c)
Expect(err).NotTo(HaveOccurred(), "while waiting for fs resize to finish")
pvcConditions := pvc.Status.Conditions
Expect(len(pvcConditions)).To(Equal(0), "pvc should not have conditions")
})
})
func waitForDeploymentToRecreatePod(client clientset.Interface, deployment *apps.Deployment) (v1.Pod, error) {
var runningPod v1.Pod
waitErr := wait.PollImmediate(10*time.Second, 5*time.Minute, func() (bool, error) {
podList, err := framework.GetPodsForDeployment(client, deployment)
for _, pod := range podList.Items {
switch pod.Status.Phase {
case v1.PodRunning:
runningPod = pod
return true, nil
case v1.PodFailed, v1.PodSucceeded:
return false, conditions.ErrPodCompleted
}
return false, nil
}
return false, err
})
return runningPod, waitErr
}
此处可能存在不合适展示的内容,页面不予展示。您可通过相关编辑功能自查并修改。
如您确认内容无涉及 不当用语 / 纯广告导流 / 暴力 / 低俗色情 / 侵权 / 盗版 / 虚假 / 无价值内容或违法国家有关法律法规的内容,可点击提交进行申诉,我们将尽快为您处理。