[OpenAFS] backups..

Corey Kovacs ckovacs@DEPAUW.EDU
Wed, 17 Oct 2001 16:28:22 -0500


I've been wrestling with the backup system for a while, mostly just
getting
the information about the tape unit hand me down I have. So far I have
gotten
the afs backup system to dump all volumes to the tape with success
reported.
The problems I am having now may or may not be related to the afs backup

system but I'll ask here just in case.

After I start the backup with

backup> dump all /full 0

I get a warning from butc in the form of...

Warning: Tape error while reading label (will proceed with dump)
        butm: no label on tape: Cannot allocate memory

then after the dump completes, I get

Warning: can't close tape
        butm: erro during tape close: Invalid argument
Updating database
Updating database - done
all.full (DumpId 1003351218): Finished. 91 volumes dumped



It seems from reading the docs that the naming issue is really not an
issue since butc
will name the tape automatically (not the perm name I know).

The thing that has got me stumped is the tape closing...

here is some more info...

 cat /proc/scsi/scsi
Attached devices:
Host: scsi0 Channel: 00 Id: 04 Lun: 00
Vendor: ARCHIVE  Model: Python 28454-XXX Rev: 4.44
Type:   Sequential-Access                ANSI SCSI revision: 02

 mt status
SCSI 2 tape drive:
File number=93, block number=0, partition=0.
Tape block size 0 bytes. Density code 0x8c (EXB-8505 compressed).
Soft error count since last status=0
General status bits on (81010000):
 EOF ONLINE IM_REP_EN

is there something else I have to do? Seems that file number=93 is
indicating that all 91
volumes plus file markers are being written but I still don't like the
error...

the version of OpenAFS I am using is 1.1.1 on a redhat 7.1 smp box
kernel 2.4.3-12.

Any info, or pointers to information on this would be great. Thanx

Corey