ceph-csi/docs/ceph-mount-corruption.md

Ceph mount corruption detection and recover

ceph-fuse: detection of corrupted mounts and their recovery

Mounts managed by ceph-fuse may get corrupted by e.g. the ceph-fuse process exiting abruptly, or its parent Node Plugin container being terminated, taking down its child processes with it.

This may manifest in concerned workloads like so:

# mount | grep fuse
ceph-fuse on /cephfs-share type fuse.ceph-fuse (rw,nosuid,nodev,relatime,user_id=0,group_id=0,allow_other)
# ls /cephfs-share
ls: /cephfs-share: Socket not connected

or,

# stat /home/kubelet/pods/ae344b80-3b07-4589-b1a1-ca75fa9debf2/volumes/kubernetes.io~csi/pvc-ec69de59-7823-4840-8eee-544f8261fef0/mount: transport endpoint is not connected

This feature allows CSI CephFS plugin to be able to detect if a ceph-fuse mount is corrupted during the volume publishing phase, and will attempt to recover it for the newly scheduled pod. Pods that already reside on a node whose ceph-fuse mountpoints were broken may still need to be restarted, however.

ceph-fuse corruption detection

A mountpoint is deemed corrupted if stat()-ing it returns one of the following errors:

• ENOTCONN
• ESTALE
• EIO
• EACCES
• EHOSTDOWN

ceph-fuse recovery

Once a mountpoint corruption is detected, its recovery is performed by remounting the volume associated with it.

Recovery is attempted only if /csi/mountinfo directory is made available to CSI CephFS plugin (available by default in the Helm chart and Kubernetes manifests).

kernel client: detection of corrupted mounts and their recovery

Mounts managed by ceph-kernel may get corrupted e.g. if your network connection is disrupted for a long enough time, the client will be forcibly disconnected from the system. More details can be found here

The above case may manifest in concerned workloads like so:

# mount | grep ceph
10.102.104.172:6789:/volumes/csi/csi-vol-7fed1ce7-97cf-43ef-9b84-2a49ab992515/d61be75e-74ae-428c-a5d1-48f79d1d3c8c on /var/lib/kubelet/plugins/kubernetes.io/csi/cephfs.csi.ceph.com/bc0146ec2b5d9a9db62e698abbe0adcae19c0e01f5cf15d3d593ed33c7bc1a8d/globalmount type ceph (rw,relatime,name=csi-cephfs-node,secret=<hidden>,fsid=00000000-0000-0000-0000-000000000000,acl,mds_namespace=myfs,_netdev)
10.102.104.172:6789:/volumes/csi/csi-vol-7fed1ce7-97cf-43ef-9b84-2a49ab992515/d61be75e-74ae-428c-a5d1-48f79d1d3c8c on /var/lib/kubelet/pods/8087df68-9756-4f38-86ef-6c81e1075607/volumes/kubernetes.io~csi/pvc-15e63d0a-77de-4886-8d0f-516f9fecbeb4/mount type ceph (rw,relatime,name=csi-cephfs-node,secret=<hidden>,fsid=00000000-0000-0000-0000-000000000000,acl,mds_namespace=myfs,_netdev)# ls /cephfs-share

sh-4.4# ls /var/lib/kubelet/plugins/kubernetes.io/csi/cephfs.csi.ceph.com/bc0146ec2b5d9a9db62e698abbe0adcae19c0e01f5cf15d3d593ed33c7bc1a8d/globalmount
ls: cannot access '/var/lib/kubelet/plugins/kubernetes.io/csi/cephfs.csi.ceph.com/bc0146ec2b5d9a9db62e698abbe0adcae19c0e01f5cf15d3d593ed33c7bc1a8d/globalmount': Permission denied

kernel client corruption detection

A mountpoint is deemed corrupted if stat()-ing it returns one of the following errors:

• ENOTCONN
• ESTALE
• EIO
• EACCES
• EHOSTDOWN

More details about the error codes can be found here

For such mounts, The CephCSI nodeplugin returns volume_condition as abnormal for NodeGetVolumeStats RPC call.

kernel client recovery

Once a mountpoint corruption is detected, Below are the two methods to recover from it.

• Reboot the node where the abnormal volume behavior is observed.
• Scale down all the applications using the CephFS PVC on the node where abnormal mounts are present. Once all the applications are deleted, scale up the application to remount the CephFS PVC to application pods.