This morning, I decided to try out an amverify on a full fump of all my servers.
I got: Loading 1 slot... Using device /dev/tape Waiting for device to go ready... Rewinding... Processing label... Volume ArchSet1-001, Date 20050602 Rewinding... Checked produc-new._disk11.20050602.0 Checked gobelet._disk1.20050602.0 Checked produc._disk10.20050602.0 Checked lnx-que-amanda._.20050602.0 Checked produc-new._disk1.20050602.0 ** Error detected (produc._disk3.20050602.0) amrestore: WARNING: not at start of tape, file numbers will be offset amrestore: 0: restoring produc._disk3.20050602.0 amrestore: read error: Input/output error gzip: stdin: unexpected end of file /bin/gtar: Read 5549 bytes from - /bin/gtar: Unexpected EOF in archive /bin/gtar: Error is not recoverable: exiting now 64+0 in 64+0 out Meanwhile, my amcheck cron job started. Unfortunately, I was away from my desk when the error occured, so I don't know if it happened at the time the amcheck job started. Could that amcheck job have cause the I/O error ? This is a 5 slot DLT changer using chg-zd-mtx If so, I wonder why amanda does not use some sort of locking mechanism while the tape is in use, to prevent such errors.
