A few short days ago I wrote:
>Greetings Managers.
>I've received strange warnings from ufsdump backup about one of the striped
>disks. That disk contains vital data and and any loss would put us
>up the proverbial creek.
>GENERAL DESCRIPTION:
>We do backups using Budtools v 4.4, that in turn uses ufsdump to write
>to the tapes. A significant number of our disks is accessed through
>the Solstice DiskSuite, v 4.0, patched up with 102580-07.
>The system is Sparc 1000E, running under Solaris 2.4, patched up
>to the recommended level.
>Disk description:
>3 Seagate ST5150W disks, 4.2 GB each, striped to form a 12 GB meta-disk.
>96% full at the time <=> 600 Mb was free.
>
>WARNINGS:
>========================================================================
>11/16/96 23:13:27> Calling Media Manager to verify the current volume.
>11/16/96 23:13:27> Using volume 'WEEKLY_961116_12' with append mode.
> .......................
> .......................
>11/16/96 23:22:19> ------ W A R N I N G -------
>
>11/16/96 23:22:19> Unexpected log entries
>11/16/96 23:22:19> DUMP: Warning - block 2050083262 is beyond the end of
>`/dev/md/rdsk/d13'
>.........and so on for 485 more lines, listing blocks
>========================================================================
>
>WHAT I CHECKED:
>I've run metatools to check for any disk errors - no problems.
>I've run fsck -n (I can't take that disk offline right now) - no problems.
>I've run format to check for any overlapping partitions - no such thing.
>I've re-calculated disk sizes - everything adds up.
>I've checked the relevant logs (/var/adm/message) files - nothing wrong.
>THE QUESTION:
>Has anyone seen something of this kind ?? Any hints, suggestions ??
>Thanks in advance,
>I'll gladly summarise....
I've received no answers, however the problem sorted itself out...... sort
of.....
when the disks in question quietly dropped dead one day after I posted my
query.
They are housed in the shonky third party cabinets and after I shutdown the
server, reset
the SCSI plugs in the cabinet and rebooted the disks kicked back in without
any complaints.
I can't be sure but I suspect that the warnings were caused by some
flakyiness
in the SCSI caused by a loose plug (Why no SCSI errors in logs though ??!!).
The conclusion is:
BUY DECENT DISK ENCLOSURES, THEY'LL PAY FOR
THEMSELVES IN THE END.
Chris Wozniak
System Administrator
WAPET Exploration
kaw@wapet.com.au
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:11:16 CDT