MC Service Guard ....

[HutchAdmin]

HI,

I am running MCService Guard Two node cluster attach with EMC DMX 1000 Storage. I want to increase the filesystem size under the cluster VG. There are free PE in Cluster VG and on-line JFS is installed on the node.

I have tried to increase the size of required filesystem while cluster is up and running. but system is not allowing me to do so.

Can anyone tell me , Is it possible to increase the file system size under cluster VG while cluster is up and running. If yes, pls help me to provide the process.

Thanks

NV Pathak

P.S. this trhead has been moved from HP-UX > HP-UX Technical Documentation to HP-UX > Serviceguard - Hp Forums moderator

[G. Vrijhoeven]

Hi Pathak,

I assume you have onlinejfs and do not have nested mountpoints. ( /mountpoint/submountpoint) if so onlinejfs does not work for /mountpoint.

Use the lvextend -L (size in mb) /dev/vg/lvol command to expand the lvol and fsadm -F vxfs -b M /mountpoint

Gideon

[G. Vrijhoeven]

Hi,

If the lvextend does not work it can be becouse you created the lvols with a phisicalvolumegroup. You could check /etc/lvmpvg and make sure the Phisical volume group needed is defined. ( lvdisplay -v /dev/vg/lvol

Gideon

[HutchAdmin]

Hi Gideon,

Thanks for the reply.

Does same applies for extending VG. I am performing following step while cluster is up and running on the primary node.

1) pvcreate disk
2) vgextend vgname diskno
3) exporting the VG and creating map files.
4) copy this map file to the adaptive node.
5) re-import the VG on adoptive node.

Is it OK...

Pathak

[G. Vrijhoeven]

Hi,

If you extend a lvol you do not have to distribute the info over all nodes. Since the vginfo does not change.
If you add/delete a lvol or extend a volume group with a disk you need to distribute. The steps you provide must work fine as long as you devicefiles/instance numbers are the same on all servers. ( c) If not you have to adjust the device files before you import.

Gideon

[G. Vrijhoeven]

Hi,

The export can be done with the -p option. What you can do is activate the vg on the alternative node read only and run an vgcfgbackup and deactivate again. ( vgchange -a r vgname && vgcfgbackup vgname & vgchange -a n vgname)

Gideon

Thanks for the points in the fsck question :-)