| Previous | Next |
| EMSGSIZE | ENETDOWN |
ESTALE
What a stale handle tells you
ESTALE is most familiar on NFS. The client holds a file handle that once identified an object on the server, but the server can no longer validate that mapping. This can follow server-side replacement, deletion and recreation, export changes, failover, snapshot rollback, or other namespace and storage events that invalidate the identity cached by the client.
The safe response is to discard the handle and resolve the object again through the current namespace. Repeating the exact operation through the same stale descriptor is not a recovery strategy. A remount may be required for a damaged client mount, but it should not be the only diagnosis: blindly remounting can hide a server migration, storage, or deployment workflow that will recreate the problem.
How to investigate
- Correlate the first failure with NFS server, cluster, export, failover, and storage-management events.
- Determine whether the application retained a directory or file descriptor across a release or namespace change.
- Close stale descriptors, re-resolve the path, and retry only an idempotent operation or one protected by an application acknowledgement.
- For long-lived clients, design an explicit recovery path rather than assuming object identity survives every server-side change.
References
Looking for a different code? Search another status or error code.