[BUG] delete mountpods 后出现多个mountpod ,应用pod写入bad file

[YunhuiChen]

What happened:
k delete pods juicefs-chaos-k8s-001-pvc-051d0a21-0983-465b-ac67-ddb79dbd15e6-uusphw
k delete pods juicefs-chaos-k8s-001-pvc-1ad6fe2f-ba3c-413b-aaef-e86353bf40d6-ueyelc
后出现了4个新的mount pod ：
root@chaos-k8s-001:~/chenyunhui# k get pods |grep pvc
juicefs-chaos-k8s-001-pvc-051d0a21-0983-465b-ac67-ddb79dbd15e6-vxmong 1/1 Running 0 2m38s
juicefs-chaos-k8s-001-pvc-1ad6fe2f-ba3c-413b-aaef-e86353bf40d6-mfknus 1/1 Running 0 57s
juicefs-chaos-k8s-001-pvc-1ad6fe2f-ba3c-413b-aaef-e86353bf40d6-tvobac 1/1 Running 0 57s
juicefs-chaos-k8s-001-pvc-1ad6fe2f-ba3c-413b-aaef-e86353bf40d6-wuwaoo 1/1 Running 0 57s
进入应用pod后：
root@dynamic-ee-0:/# echo 1 > /data/2
bash: echo: write error: Bad file descriptor

E0210 10:23:08.870513       7 pod_driver.go:733] "bind mount error" err="function timeout" logger="pod-driver" podName="juicefs-chaos-k8s-001-pvc-051d0a21-0983-465b-ac67-ddb79dbd15e6-vxmong"
E0210 10:23:08.870588       7 reconciler.go:145] "Driver check pod error, will retry" err="exec cmd: mount -o bind /jfs/pvc-051d0a21-0983-465b-ac67-ddb79dbd15e6-uusphw/pvc-051d0a21-0983-465b-ac67-ddb79dbd15e6 /var/lib/kubelet/pods/43d5de36-65af-48a4-8a22-773ac08243b0/volumes/kubernetes.io~csi/pvc-051d0a21-0983-465b-ac67-ddb79dbd15e6/mount err:function timeout" logger="reconciler-controller" name="juicefs-chaos-k8s-001-pvc-051d0a21-0983-465b-ac67-ddb79dbd15e6-vxmong"
I0210 10:23:13.891017       7 pod.go:260] "Mount point is ready" logger="pod-driver" podName="juicefs-chaos-k8s-001-pvc-051d0a21-0983-465b-ac67-ddb79dbd15e6-vxmong"
I0210 10:23:14.192817       7 pod.go:260] "Mount point is ready" logger="pod-driver" podName="juicefs-chaos-k8s-001-juicefs-console-pv-yundlg"
I0210 10:23:14.893133       7 pod_driver.go:708] "umount pod target before recover" logger="pod-driver" podName="juicefs-chaos-k8s-001-pvc-051d0a21-0983-465b-ac67-ddb79dbd15e6-vxmong" target="/var/lib/kubelet/pods/43d5de36-65af-48a4-8a22-773ac08243b0/volumes/kubernetes.io~csi/pvc-051d0a21-0983-465b-ac67-ddb79dbd15e6/mount" remain mount count=5
I0210 10:23:15.894852       7 pod_driver.go:726] "recover volPath" logger="pod-driver" podName="juicefs-chaos-k8s-001-pvc-051d0a21-0983-465b-ac67-ddb79dbd15e6-vxmong" target="/var/lib/kubelet/pods/43d5de36-65af-48a4-8a22-773ac08243b0/volumes/kubernetes.io~csi/pvc-051d0a21-0983-465b-ac67-ddb79dbd15e6/mount" mountPath="/jfs/pvc-051d0a21-0983-465b-ac67-ddb79dbd15e6-uusphw/pvc-051d0a21-0983-465b-ac67-ddb79dbd15e6"

What you expected to happen:

How to reproduce it (as minimally and precisely as possible):

Anything else we need to know?

Environment:

• JuiceFS CSI Driver version (which image tag did your CSI Driver use):
• Kubernetes version (e.g. kubectl version):
• Object storage (cloud provider and region):
• Metadata engine info (version, cloud provider managed or self maintained):
• Network connectivity (JuiceFS to metadata engine, JuiceFS to object storage):
• Others: