Applicable versions

All Longhorn versions.

Symptoms

The pod using a longhorn volume with an ext4 filesystem stays in container Creating with errors in the log.

  Warning  FailedMount             30s (x7 over 63s)  kubelet                  MountVolume.SetUp failed for volume "pvc-bb8582d5-eaa4-479a-b4bf-328d1ef1785d" : rpc error: code = Internal desc = 'fsck' found errors on device /dev/longhorn/pvc-bb8582d5-eaa4-479a-b4bf-328d1ef1785d but could not correct them: fsck from util-linux 2.31.1
ext2fs_check_if_mount: Can't check if filesystem is mounted due to missing mtab file while determining whether /dev/longhorn/pvc-bb8582d5-eaa4-479a-b4bf-328d1ef1785d is mounted.
/dev/longhorn/pvc-bb8582d5-eaa4-479a-b4bf-328d1ef1785d contains a file system with errors, check forced.
/dev/longhorn/pvc-bb8582d5-eaa4-479a-b4bf-328d1ef1785d: Inodes that were part of a corrupted orphan linked list found.  

/dev/longhorn/pvc-bb8582d5-eaa4-479a-b4bf-328d1ef1785d: UNEXPECTED INCONSISTENCY; RUN fsck MANUALLY.
  (i.e., without -a or -p options)

Reason

Longhorn cannot remount the volume when the Longhorn volume has a corrupted filesystem. The workload then fails to restart as a result of this.

Longhorn cannot fix this automatically. You will need to resolve this manually when this happens.

Solution

For most Linux distribution versions

1. Search for error indicators:
   • Check if the volume is in an error state from the Longhorn UI.
   • Check Longhorn manager pods log for system corruption error messages.
   • If the volume is not in an error state then the file system inside Longhorn volume may be corrupted by an external reason.
2. Scale down the workload.
3. Attach the volume to any node from the UI.

Warning When a file system check tool fixes errors, it modifies the filesystem metadata and brings the filesystem to a consistent state. However, an incorrect fix might lead to unexpected data loss or more serious filesystem corruption. To mitigate the potential risk, we highly suggest that users take a snapshot or a backup of the corrupted filesystem before attempting any fix. In case of an accident, users can recover the volume.

4. SSH into the node.
5. Find the block device corresponding to the Longhorn volume under /dev/longhorn/<volume-name>.
6. Use a filesystem check tool to repair the filesystem. For example:
   • Fix an ext4 filesystem using fsck.
   • Fix an xfs filesystem using xfs_repair.
7. On the Longhorn UI, detach the volume.
8. Scale up the workload.

For some older Linux distribution versions and Longhorn volumes with ext4 filesystems

In the CSI flow, the Longhorn CSI plugin creates a file system on a new volume using the make2fs utility (command: mkfs.ext4) built into its container. The e2fsck utility (command: fsck.ext4) available in some older Linux distributions may not support all features this file system is created with, resulting in the following error:

-> fsck.ext4 /dev/longhorn/pvc-c7152ef5-55c7-43ce-a35e-dac69d2be591 
e2fsck 1.42.9 (28-Dec-2013)
/dev/longhorn/pvc-c7152ef5-55c7-43ce-a35e-dac69d2be591 has unsupported feature(s): metadata_csum
e2fsck: Get a newer version of e2fsck!

If possible, upgrade your e2fsprogs (Ext2/3/4 Filesystem Utilities) to a later version. If upgrading is not possible (for example, you are running CentOS 7 or RHEL 7), you can access attached Longhorn volumes using the updated e2fsck that is built into the instance-manager or instance-manager-e container.

1. Search for error indicators:
   • Check if the volume is in an error state from the Longhorn UI.
   • Check Longhorn manager pods log for system corruption error messages.
   • If the volume is not in an error state then the file system inside Longhorn volume may be corrupted by an external reason.
2. Scale down the workload.
3. Attach the volume to any node from the UI.

Warning When a file system check tool fixes errors, it modifies the filesystem metadata and brings the filesystem to a consistent state. However, an incorrect fix might lead to unexpected data loss or more serious filesystem corruption. To mitigate the potential risk, we highly suggest that users take a snapshot or a backup of the corrupted filesystem before attempting any fix. In case of an accident, users can recover the volume.

4. Open a shell inside the instance-manager or instance-manager-e pod running on the node that the volume is attached to:
   kubectl exec -it -n longhorn-system instance-manager-<additional-characters> -- bash
5. Find the block device corresponding to the Longhorn volume under /dev/longhorn/<volume-name>.
6. Use a filesystem check tool to repair the filesystem. For example,
   • Fix an ext4 filesystem using fsck.
   • Fix an xfs filesystem using xfs_repair.
7. On the Longhorn UI, detach the volume.
8. Scale up the workload.

Example output using Longhorn v1.4.0 (with e2fsprogs v1.46.4) and CentOS 7.9 (with e2fsprogs v1.42.9) :

-> kl exec -it instance-manager-e-545c3360290f259fb0fe5638303b8f9a bash

instance-manager-e-545c3360290f259fb0fe5638303b8f9a:/ # fsck.ext4 /dev/longhorn/pvc-c7152ef5-55c7-43ce-a35e-dac69d2be591 
e2fsck 1.46.4 (18-Aug-2021)
/dev/longhorn/pvc-c7152ef5-55c7-43ce-a35e-dac69d2be591: clean, 11/131072 files, 26156/524288 blocks