Hi everyone,
I have a BeeGFS environment running version 8.2.2, and with the main services in the same host (mgmt, meta, and storage), and recently had a disk failure in one of my storage disks, causing failures for the BeeGFS clients, but I was able to fix the issue and restore the connection. Because of that, I ran, from a client with root permissions, the beegfs-fsck first on readOnly and then without the --readOnly flag to fix, in online mode, the issues caused by the disk failure.
Now, every time I try to run `beegfs-fsck --readOnly` from the client, I'm getting the following message:
```
# beegfs-fsck --checkfs --readOnly --databasePath=/mnt/data/backup
BeeGFS File System Check Version : 8.2.2
----
--------------------------------------------------------------------
Started BeeGFS fsck in forward check mode [Mon Mar 9 15:32:07 2026]
Log will be written to /var/log/beegfs-fsck.log
Database will be saved in /mnt/data/backup
--------------------------------------------------------------------
Step 1: Check reachability of nodes: Finished
The database path already exists: /mnt/data/backup/fsckdb
If you continue now any existing database files in that path will be deleted.
Do you want to continue? (Y/N)
y
-----
Waiting for metadata servers to start modification logging. This might take some time.
-----
Step 3: Gather data from nodes:
Fetched data > Directory entries: 1000 | Inodes: 1876 | Chunks: 0
-----
Waiting for metadata servers to stop modification logging. This might take some time.
-----
Another instance of beegfs-fsck is still running or was aborted. Cannot start a new one unless --forceRestart command line switch is used. Do this only after making sure there is no other instance of beefs-fsck running.
```
And these messages from the BeeGFS server (`/var/log/message`):
```
Mar 9 15:32:10 beegfs01 beegfs-storage[8210]: Worker4-1 [FetchFsckChunkListMsgEx.cpp:22] >> Received request to start fsck although previous run is not finished. Not starting. From:
10.50.0.9:48517Mar 9 15:32:10 beegfs01 beegfs-meta[7939]: Worker1 [DiskMetadata (DirInode Deserialization)] >> Deserialization failed: expected DirInode, but got (numeric type): 3
Mar 9 15:32:11 beegfs01 beegfs-meta[7939]: Worker1 [Directory (load from xattr file)] >> Unable to deserialize metadata in file: inodes/0/0/7B48-69AEC7A1-2
Mar 9 15:32:11 beegfs01 beegfs-meta[7939]: Worker17 [DiskMetadata (DirInode Deserialization)] >> Deserialization failed: expected DirInode, but got (numeric type): 3
Mar 9 15:32:11 beegfs01 beegfs-meta[7939]: Worker17 [Directory (load from xattr file)] >> Unable to deserialize metadata in file: inodes/41/0/8638-69AED939-2
Mar 9 15:32:11 beegfs01 beegfs-meta[7939]: Worker17 [DiskMetadata (DirInode Deserialization)] >> Deserialization failed: expected DirInode, but got (numeric type): 3
Mar 9 15:32:11 beegfs01 beegfs-meta[7939]: Worker17 [Directory (load from xattr file)] >> Unable to deserialize metadata in file: inodes/41/0/15F7-69AE4AE8-2
Mar 9 15:32:11 beegfs01 beegfs-meta[7939]: Worker17 [DiskMetadata (DirInode Deserialization)] >> Deserialization failed: expected DirInode, but got (numeric type): 3
Mar 9 15:32:11 beegfs01 beegfs-meta[7939]: Worker17 [Directory (load from xattr file)] >> Unable to deserialize metadata in file: inodes/41/0/574C-69A97B6E-2
Mar 9 15:32:11 beegfs01 beegfs-meta[7939]: Worker1 [DiskMetadata (DirInode Deserialization)] >> Deserialization failed: expected DirInode, but got (numeric type): 3
Mar 9 15:32:11 beegfs01 beegfs-meta[7939]: Worker1 [Directory (load from xattr file)] >> Unable to deserialize metadata in file: inodes/0/0/9A9E-69A6B547-2
(...)
Mar 9 15:32:11 beegfs01 beegfs-meta[7939]: Worker17 [DiskMetadata (DirInode Deserialization)] >> Deserialization failed: expected DirInode, but got (numeric type): 3
Mar 9 15:32:11 beegfs01 beegfs-meta[7939]: Worker17 [Directory (load from xattr file)] >> Unable to deserialize metadata in file: inodes/41/0/71C7-69A6B547-2
Mar 9 15:32:11 beegfs01 beegfs-meta[7939]: Worker1 [DiskMetadata (DirInode Deserialization)] >> Deserialization failed: expected DirInode, but got (numeric type): 3
Mar 9 15:32:11 beegfs01 beegfs-meta[7939]: Worker1 [Directory (load from xattr file)] >> Unable to deserialize metadata in file: inodes/0/0/A51E-69A93DEB-2
Mar 9 15:33:50 beegfs01 beegfs-meta[7939]: ModificationEventFlusher [ModificationEventFlusher] >> Did not receive an ack from fsck for a FsckModificationEventMsg
```
Is it safe to force the new beegfs-fsck, or should I do something else?
Thank you in advance!