| Previous | Next |
| EINTR | ENXIO |
EIO
Why this error needs context
EIO is a broad low-level I/O result, not a diagnosis by itself. A read can return it for an underlying disk or tape failure, for a lost advisory lock on a network filesystem, or for a terminal-job-control condition. File-creation and metadata operations may use the same errno when the filesystem cannot complete an update.
Because the same code can describe a transient transport problem, a device fault, a filesystem problem, or an API-specific state, a retry policy should be based on the operation. Retrying an idempotent read after recording the context can be reasonable; blindly repeating a write can duplicate or partially apply work.
What to preserve for diagnosis
- Record the syscall, path or descriptor type, offset, requested byte count, and whether any data was already transferred.
- Check kernel, storage, transport, and filesystem logs around the same timestamp.
- For network filesystems, verify mount/server health and whether application locking is involved.
- For persistent device errors, protect data first: stop destructive retries and verify backups before repair or replacement work.
References
Looking for a different code? Search another status or error code.