LVM and VxVM

vmunix: vxfs: mesg 001: vx_nospace - /dev/vg00/lvol9 file system full (1 block extent)

 
WilliamSmith11
Super Advisor

vmunix: vxfs: mesg 001: vx_nospace - /dev/vg00/lvol9 file system full (1 block extent)

Hi
I have a quetion about the posted question.

Can a file system full do that a package of a cluster node be halted?

I had this message:
vmunix: vxfs: mesg 001: vx_nospace - /dev/vg00/lvol9 file system full (1 block extent)

CM-CMD[20687]: cmhaltpkg ora816
Apr 6 19:26:42 Denver1 cmcld: Request from node Denver2 to disable global switching for package ora816.

Apr 6 19:26:42 Denver1 cmcld: Request from node Denver2 to halt package ora816 on node Denver1

Apr 6 19:26:42 Denver1 cmcld: Executing '/etc/cmcluster/ora816/ora816.cntl stop' for package ora816, as service PKG*32769.

Apr 6 19:26:42 Denver1 CM-ora816[20702]: cmhaltserv ora816

Thank you

W.S




rperez
4 REPLIES 4
A. Clay Stephenson
Acclaimed Contributor

Re: vmunix: vxfs: mesg 001: vx_nospace - /dev/vg00/lvol9 file system full (1 block extent)

Yes. It all depends upon what the control and monitoring scripts tells the package to do. It's also possible that this full filesystem had nothing to do with the package. It would help if you explained what vg00/lvol9 was used for. What filesystem is that since it's not a typical filesystem for vg00.
If it ain't broke, I can fix that.
Mark Greene_1
Honored Contributor

Re: vmunix: vxfs: mesg 001: vx_nospace - /dev/vg00/lvol9 file system full (1 block extent)

I assume that the ora816 is Oracle, yes?

Check when your backups run. If you are doing dumps at night you may have run out of space if you are keeping more than 1 day's of dumps on disk. Also check if the redo logs are in this volume group, as they may have consumed the space.

mark
the future will be a lot like now, only later
Torsten.
Acclaimed Contributor

Re: vmunix: vxfs: mesg 001: vx_nospace - /dev/vg00/lvol9 file system full (1 block extent)

Typical LVOLs for vg00 are from 1 to 8. So LVOL9 is a custom LVOL.

No serviceguard process should write to vg00, because this cannot switched to another host.

If a process cannot write its data, its reasonable enough to cause an error - the process fail and the failure cause a package
switch.

Increase space and investigate your packages!


Hope this helps!
Regards
Torsten.

__________________________________________________
There are only 10 types of people in the world -
those who understand binary, and those who don't.

__________________________________________________
No support by private messages. Please ask the forum!

If you feel this was helpful please click the KUDOS! thumb below!   
Josiah Henline
Valued Contributor

Re: vmunix: vxfs: mesg 001: vx_nospace - /dev/vg00/lvol9 file system full (1 block extent)

You package log file, usually in /etc/cmcluster/PKGDIR/PKG.cntl.log may have better details as to why the package failed over.
If at first you don't succeed, read the man page.