dgraph-io/dgraph · error

RunRestore failed to write group id file

Error message

RunRestore failed to write group id file

What it means

After successfully flushing the restored DB into the post directory, offline restore writes a GROUP_ID file inside p<gid> identifying which group this posting directory belongs to. If x.WriteGroupIdFile fails (I/O error writing the file), the restore result is rejected because the Alpha would later fail to load the directory without the group marker.

Source

Thrown at worker/online_restore.go:609

			WithEncryptionKey(key).
			WithNamespaceOffset(x.NamespaceOffset))
		if err != nil {
			return LoadResult{Err: errors.Wrap(err, "RunRestore failed to open DB")}
		}
		defer db.Close()

		sw := db.NewStreamWriter()
		if err := sw.Prepare(); err != nil {
			return LoadResult{Err: errors.Wrap(err, "while preparing DB")}
		}
		if err := RunReducer(sw, mapDir); err != nil {
			return LoadResult{Err: errors.Wrap(err, "RunRestore failed to reduce")}
		}
		if err := sw.Flush(); err != nil {
			return LoadResult{Err: errors.Wrap(err, "while stream writer flush")}
		}
		if err := x.WriteGroupIdFile(pdir, gid); err != nil {
			return LoadResult{Err: errors.Wrap(err, "RunRestore failed to write group id file")}
		}
	}
	// TODO: Fix this return value.
	return LoadResult{Version: manifest.ValidReadTs()}
}

func buildPredsForDefaultNamespace(restorePreds []string, fromNamespace uint64) []string {
	filtered := restorePreds[:0]
	for _, pred := range restorePreds {
		ns, attr := x.ParseNamespaceAttr(pred)
		if fromNamespace == ns {
			// update namespace value to 0
			pred = x.NamespaceAttr(0, attr)
			filtered = append(filtered, pred)
		}
	}
	return filtered
}

View on GitHub (pinned to 759e242be6)

Solutions

1. Check write permissions/ownership of the target postings directory for the process user (chown/chmod).
2. Verify the filesystem is not read-only or full.
3. Ensure the p<gid> directory still exists at the end of restore (no external cleanup jobs).
4. Retry the restore; if only the group file is missing from an otherwise complete restore, it can be created manually with the group id content before starting Alpha.

Example fix

// before
$ dgraph restore -p /data/postings  # permission denied writing p1/GROUP_ID
// after
$ sudo chown -R dgraph:dgraph /data/postings
$ dgraph restore -p /data/postings

fi, err := os.Stat(dir)
if err != nil {
	return err
}
if err := syscall.Access(dir, syscall.O_RDWR); err != nil {
	return fmt.Errorf("no write permission on %s", dir)
}

if err != nil && strings.Contains(err.Error(), "failed to write group id file") {
	// fix permissions/ownership or free space, then re-run restore
}

When it happens

Trigger: RunOfflineRestore reaches x.WriteGroupIdFile(pdir, gid) and the file write fails: read-only filesystem, no write permission on the post directory, disk full, or the directory was removed/moved mid-restore.

Common situations: Restoring into a directory owned by a different user than the dgraph process; running the restore in a container whose volume switched to read-only; filesystem quota exceeded.

Related errors

• cannot get backup manifests
• error while creating debug file: %s
• Create path failed to create path %s, got error: %v
• File handler failed to create file %s
• another restore operation is already running

AI-assisted analysis of dgraph-io/dgraph@759e242be6 (2026-09-01). Data as JSON: /api/errors/9d684634633e705a. Report an issue: GitHub.