unable to recreate a corrupt or missing lvmtab file

[SSO Montgomery]

On the problem system, I attached two 9-gig drives to the wide SCSI
interface using device numbers 3 and 4. I was able to create a VG and LV
using SAM. I had to delete these and re-create them using SAM. When I did,
the machine complained that the old volume group still existed (subject of
the original call).

Mitch told me about the "vgscan -v". This was suppose to rebuild
the LVMTAB, so I could proceed.

Get the following error:
Pid 19777 received a SIGSEGV for stack growth failure.
Possible causes: insufficient memory for swap space,
or stack size exceeded maxssiz.

Does anyone have any suggestions. Am running 11.0

[harry d brown jr]

Try this:

http://mail.gnu.org/pipermail/ddd/2001-March/000654.html

Also make sure you have the latest patches as this thread shows:

http://inf33-www.informatik.unibw-muenchen.de/infothek/computer/hp-ux/Patches_1998_05_15.txt


live free or die

harry

[Ron Cornwell]

It is most likely that you are close to your limit on swap space. Do a swapinfo and see how much swap space is being used. You can free swap space by having users log off or shutting down unnecessary applications for the time being. Otherwise check the mount of data space available in maxdsiz sysdef | grep maxd

[S.K. Chan]

vgscan is probably complaining about the 2x9GB disks that you've just installed.

Look at output from ..
/etc/diskinfo /dev/rdsk/
and see if there's anything weird.

If you have any non-LVM disks or disks which are installed but not used, you can temporarily move their device file to somewhere first and run vgscan again, just to make sure those disks are not causing the problem.

[James R. Ferguson]

HI;

Stack growth failure points to 'maxdsiz'/'maxdsiz_64bit' or perhaps 'maxssiz/maxssiz_64bit' . You probably want to increase these values and regenerate your kernel. SAM will do a nice job of this for you. See this document for more information:

http://docs.hp.com/hpux/onlinedocs/os/KCparams.OverviewAll.html

It is also worthwhile to make sure that you have sufficient swap available. Check that with 'swapinfo -tam'.

Regards!

...JRF...
http://docs.hp.com/hpux/onlinedocs/os/KCparam.MaxDsiz.html

[A. Clay Stephenson]

Did you first mv /etc/lvmtab to /etc/lvmtab.old? In your case, you probably don't want the existing data to 'assist' vgscan.