package down

[Fauziah Mahdan]

Hi i just extend my logical volume which running in cluster and sharing mode.
the command is
1) halt package
2) 1) halt package
cmhaltpkg -v oracle1pkg (for 1st pckg)
cmhaltpkg -v oracle2pkg (for 2nd pckg)

2) vgchange -c n vg03
3) vgchange -a y vg03
4) lvextend -L 1000 /dev/vg03/lvol2
5) vgchange -a n vg03
6) start package
cmrunpkg –v oracle1pkg (for 1st pckg)
cmrunpkg –v oracle2pkg (for 2nd pckg)

step no 6 not successful
what shoule we do?

[RAC_1]

What is the error that it gave??

5) vgchange -a n vg03

Why? It should be vgchange -a e vg03.

[Gary L. Paveza, Jr.]

In step 2 you did a vgchange -c n vg03

You need to put the cluster mark back on the volume group:

vgchange -c y /dev/vg03

[Rita C Workman]

You could vgchange -a e
or
you could vgchange -a n
then vgchange -c y

or....if you have OnlineJFS, why did you take it down...just extend it live.

Just a thought,
Rita

[Mel Burslan]

As Gary stated, you need to mark this disk back into the cluster aware status with

vgchange -c y vg03

then

vgchange -a -n vg03

then

cmrunpkg pkg_name

Also, as Rita indicated, I am assuming, if you have invested in MC ServiceGuard, I am almost certain that you have onlineJFS installed on your system. Why not use it and save yourself a lot of trouble ?

[Pedro Cirne]

Hi Rita & Mel,

I guess this is an Oracle RAC/OPS Cluster with raw devices and VG in shared mode. In this cenário it's not possible to extend on-line, even if you have OnLine JFS.

Enjoy :)

Pedro

[Fauziah Mahdan]

Hi all,
Thanks
for the respond.
I manage to extend the lv..after log a call to HP..