From: Eiler, James A. (James.Eiler@alcoa.com)
Date: Wed Nov 20 2002 - 03:08:29 EST
Original question is at the bottom...
Basic problem seen by most everyone who replied was an issue with the file
system.
At the time the system panic'ed, another "user" on the system was mounting
some partitions from an old disk. I'm not sure exactly what they were
doing, but the problem has not re-occurred...
Thanks to:
Michael James Bradford
Kjell Andresen
Dr Thomas.Blinn
Steve J. Dyer
Selden E Ball Jr
scott@zk3.dec.com
Jason Orendorf
Some selected replies:
------
this panic is normally associated with UFS root filesystem corruption
(or corruption of another UFS filesystem). Best bet is to:
1. Boot OS CD
2. Run 'fsck -oy /dev/rdisk/dsk??' on all UFS filesystems you have.
3. Boot the usual boot disk.
------
>From within kdbx, "p *pmsgbuf" will display output destined for the
console. This should include the identity of the corrupted filesystem.
The "ialloc: dup alloc" message indicates an inconsistency between an
inode map and an inode block. Specifically, the inode map indicated a
free inode, but the inode block showed that particular inode to be in use.
A common scenario for causing a dup alloc on the root filesys is to bypass
the fsck of root in the boot sequence. This is easily done by booting
into single-user mode, mounting the root filesystem read-write (mount -u
/), then later going into multi-user mode. The bcheckrc performed when
transitioning into multi-user mode performs a "fsck -p", which bypasses
writeable filesystems.
------
That is a file system panic. I've seen it happen when I had problems
doing I/O to some of my SCSI devices. You need to look in the system's
messages file (/var/adm/messages) for any interesting messages that
may have been displayed just before the panic, you need to look at the
error log (with uerf, dia, or ca), and you need to go into the crash
dump directory (/var/adm/crash) and look for the crash-data file and
see if there are any interesting things in the stored messages buffer
from before the panic. kdbx won't tell you anything useful unless you
know what to ask it to show, most of the interesting data should be
in the crash-data file.
------
try "uerf -R -o full |more" as well
------
Original Posting:
Hello,
Just installed version 5.1A, PK3 yesterday. This morning, the system
panic'ed.
uerf tells me this:
----- EVENT INFORMATION -----
EVENT CLASS ERROR EVENT
OS EVENT TYPE 302. PANIC
SEQUENCE NUMBER 75.
OPERATING SYSTEM DEC OSF/1
OCCURRED/LOGGED ON Tue Nov 19 08:54:33 2002
OCCURRED ON SYSTEM kofhm1
SYSTEM ID x0007001E
SYSTYPE x0000001A
MESSAGE panic (cpu 0): ialloc: dup alloc
kdbx tells me this:
# kdbx -k vmunix.0 vmcore.0
dbx version 5.1
Type 'help' for help.
stopped at [boot:2644 ,0xfffffc00004c9414] Source not available
warning: Files compiled -g3: parameter values probably wrong
Any suggestions on what else I could look for?
Thanks!
Jim
This archive was generated by hypermail 2.1.7 : Sat Apr 12 2008 - 10:49:00 EDT