Change Dump Destination Device?

[Brent Hardesty]

I have read many of the threads discussing dump configuration and primary swap space, but still remain a bit confussed. Currently, on our batch K370, 10.2 server, the dump is disabled. When I run dmesg this is what it reports:

WARNING: Logical volume for Dump expected but not found.
Swap device table: (start & size given in 512-byte blocks)
entry 0 - major is 64, minor is 0x2; start = 0, size = 2097152
Defaulting dump to primary swap
WARNING: Could not find HPA for dump device 0xffffffff
WARNING: Could not use primary swap for dump. Dump is disabled.

Is this telling me that dump has not been assigned to a logical volume?

I have 1.5 Gig of RAM, but in my primary swap we only have 1G configured. /var/adm/crash also has 1G configured. Savecore is set to 1.

Swapinfo -tam:
Mb Mb Mb PCT START/ Mb
TYPE AVAIL USED FREE USED LIMIT RESERVE PRI NAME
dev 1024 0 1024 0% 0 - 1 /dev/vg00/lvol2
dev 2000 0 2000 0% 0 - 0 /dev/vg02/lvswap1
dev 1936 0 1936 0% 65472 - 0 /dev/vg01/lvswap1
reserve - 2392 -2392
memory 1097 129 968 12%
total 6057 2521 3536 42% - 0 -

lvlnboot -v:
Boot Definitions for Volume Group /dev/vg00:
Physical Volumes belonging in Root Volume Group:
/dev/dsk/c0t6d0 (10/0.6.0) -- Boot Disk
/dev/dsk/c0t4d0 (10/0.4.0)
No Boot Logical Volume configured
Root: lvol3 on: /dev/dsk/c0t6d0
Swap: lvol2 on: /dev/dsk/c0t6d0
Dump: lvol2 on: /dev/dsk/c0t6d0, 0

Current path "/dev/dsk/c9t2d0" is an alternate link, skip.
Current path "/dev/dsk/c9t2d1" is an alternate link, skip.
lvlnboot: Volume group not activated.
Cannot display volume group "/dev/vglock".

Sorry for the long message...just wanted to throw out all info that you may want to see.
I guess I am trying to figure out where I need to go from here. I would like to get dump set up so I can obtain core dump files when my system crashes.

thanks for any help that you can provide....
Brent

[Brent Hardesty]

I think I have figured it out and have configured new dump on vg00. Everything looks good, will test next reboot.

thx

[Steffi Jones_1]

Hello,

you can try to run the following command:

lvlnboot -d lvol2 vg00

Assuming that you have lvol2 for primary swap.

Have a look at the manpage for lvlnboot

Let's keep fingers crossed for the next reboot :-)

Steffi Jones

[Bill McNAMARA]

After the system has rebooted following the crash, the system determines that a dump is present. By default the savecore process is then run to copy the dump into the filesystem under the /var/adm/crash directory (or another location, as configured in the /etc/rc.config.d/savecore file).

The following explains how to add dedicated dump space in the kernel
First, obtain the hardware address of disk:
# ioscan -fnkC disk
Class I H/W Path Driver S/W State H/W Type Description
==================================================================
disk 0 12/4.5.0 sdisk CLAIMED DEVICE SEAGATE ST34371W
/dev/dsk/c0t5d0 /dev/rdsk/c0t5d0
disk 1 12/4.8.0 sdisk CLAIMED DEVICE SEAGATE ST34371W
/dev/dsk/c0t8d0 /dev/rdsk/c0t8d0
disk 3 8/16/5.2.0 sdisk CLAIMED DEVICE TOSHIBA CD-ROM XM-5701T
/dev/dsk/c2t2d0 /dev/rdsk/c2t2d0
ianka:/ #

Then regenerate the kernel:
# cd /stand/build
# /usr/sysadm/lbin/get_sysfile /stand/vmunix > system
# vi system
. . .
dump on 12/4.8.0
. . .
# mk_kernel -o ../vmunix -s system
# reboot

After a panic/TOC there is no need to save the crash dump immediately as the dump area is dedicated so will not be over written by swap. The savecore command is then typically run manually to a specified directory with enough space (1/3 to 1/2 size of RAM) or directly to tape.

Later,
Bill