On Tue, 25 Sep 2001, John R. Jackson wrote:
> Ummm, application level software (like Amanda) cannot *cause* hardware
> errors. All it does are simple read and fsf calls.
Right. But the way in which it does is different from mt/dd and obviously
causes some adverse behavior to make the kernel produce those messages either
directly or indirectly.
> >However, I was able to restore with mt/dd/restore.
>
> Did you use "mt rewind/mt fsf" before amrestore/amrecover? What happens
> if you do?
I did. I also tried positioning it at various blocks.
> What did the amrestore output look like? Or, what's in amidxtaped*debug
> on the client when you run amrecover?
# mt status
SCSI 2 tape drive:
File number=0, block number=0, partition=0.
Tape block size 0 bytes. Density code 0x80 (unknown to this mt).
Soft error count since last status=0
General status bits on (41010000):
BOT ONLINE IM_REP_EN
# mt rewind
# mt tell
At block 0.
# amrecover Daily -s tapeserver.com -t tapeserver.com
AMRECOVER Version 2.4.2p2. Contacting server on tapeserver.com ...
220 tapeserver AMANDA index server (2.4.2p2) ready.
200 Access OK
Setting restore date to today (2001-09-26)
200 Working date set to 2001-09-26.
200 Config set to Daily.
200 Dump host set to tapeserver.com.
Can't determine disk and mount point from $CWD
amrecover> sethost hostname.net
200 Dump host set to hostname.net.
amrecover> setdisk hda9
Scanning /tmp/backups...
200 Disk set to hda9.
amrecover> setdate ---26
200 Working date set to 2001-09-26.
amrecover>
amrecover> cd vs
/vs
amrecover> cd customer.com
/vs/customer.com
amrecover> cd var
/vs/customer.com/var
amrecover> cd spool
/vs/customer.com/var/spool
amrecover> cd mail
/vs/customer.com/var/spool/mail
amrecover> add joeuser
Added /vs/customer.com/var/spool/mail/joeuser
amrecover> list
TAPE Daily-004 LEVEL 1 DATE 2001-09-26
/vs/customer.com/var/spool/mail/joeuser
amrecover> extract
Extracting files using tape drive /dev/tape on host tapeserver.com.
The following tapes are needed: Daily-004
Restoring files into directory /tmp/recover
Continue? [Y/n]:
Load tape Daily-004 now
Continue? [Y/n]:
EOF, check amidxtaped.debug file on tapeserver.com.
amrecover: short block 0 bytes
UNKNOWN file
amrecover: Can't read file header
extract_list - child returned non-zero status: 1
Continue? [Y/n]:
amrecover>
amidxtaped.debug:
> SECURITY USER root
bsd security: remote host tapeserver.com user root local user operator
amandahosts security check passed
> 6
amrestore_nargs=6
> -h
> -p
> /dev/tape
> hostname.net
> ^hda9$
> 20010926
Ready to execv amrestore with:
path = /usr/sbin/amrestore
argv[0] = "amrestore"
argv[1] = "-h"
argv[2] = "-p"
argv[3] = "/dev/tape"
argv[4] = "hostname.net"
argv[5] = "^hda9$"
argv[6] = "20010926"
amrestore: 0: skipping start of tape: date 20010926 label Daily-004
amrestore: error reading file header: Input/output error
amidxtaped: amrestore terminated normally with status: 2
Rewinding tape: done
amidxtaped: pid 4985 finish time Wed Sep 26 16:26:09 2001
kernel reports:
st0: Error with sense data: Info fld=0x8000, Current st09:00: sense key Medium Error
Additional sense indicates Recorded entity not found
Come to think of it, the only change I can recall from when we originally set
this up (and worked) is the firmware on the ecrix was upgraded. But I still
don't understand why mt/dd work and amrecover/amrestore do not.
--
Rob
