Originally posted by: grukrz1

the VG is NOT enhanced concurrent

when I would run "chvg -t 3 sharedvg" on active node how to propagate this VG change to the other cluster nodes (passive ones)?

anyone tested/did it on running cluster? I need to extendvg with biger disk and the factor has to be increased.
Kind Regards,
K.
#AIX-Forum

Originally posted by: MichaelAM

Unfortunately I dont have a cluster to verify right now.

My best guess at this moment is that there will be a CSPOC dialog that you can use to make the change to the volume group - also the addition of the new disk. CSPOC will ensure that the changes are updated on the other nodes ODM.

And even if you dont use CSPOC HACMP would use either "importvg -L hdiskXX" or the so-called lazy update (exportvg vgAAA, importvg -y vgAAA hdiskXX), and the changes will be known on the other node. However, fallover takes a bit longer.
#AIX-Forum

Originally posted by: grukrz1

no, there are no "chvg" substitutes in C-SPOC (that was what I checked first)

anyhow, I did test with a sharedvg, and there was nothing to sync after I run "chvg -t 3 sharedvg".

What I did step by step to test it was:

node1:
chvg -t 3 testsharedvg
varyoffvg testsharedvg

node2:
varyonvg testsharedvg
lsvg testsharedvg <--- I see "max pp per pv" up_to_date without any extra action

btw. LVM in C-SPOC supports (can be found in /usr/sbin/cluster/cspoc/smitlvm script):

1. $1 = '-1' cl_chlv
2. '-2' cl_chfs
3. '-3' (namehdr for chlv)
4. '-4' (namehdr for chfs)
5. '-5' (namehdr) returns RGNAME:NAME
6. '-6' cl_lslv
7. '-7' cl_chlv dialog
8. '-8' cl_rmlv
9. '-9' cl_rmfs
10. '-10' cl_chfs
11. '-11' cl_lspvsmit
12. '-12' cl_extendvg
13. '-13' cl_reducevg
14. '-14' cl_importvg
15. '-15' cl_mirrorvg
16. '-16' cl_unmirrorvg
17. '-17' cl_mklv
18. '-18' cl_extendlv
19. '-19' cl_mklvcopy
20. '-20' cl_rmlvcopy
21. '-21' cl_syncvg
22. '-22' (namehdr for disks) returns NODE:HDISK

NAMES

1. '-23' cl_lsfreedisks

so, no "chvg".

Kind regards,
K.

#AIX-Forum

Originally posted by: unixgrl

You should be able to do this live by breaking the reserve on the active node
varyonvg -b <diskgroup> : on active node

Then exportvg/importvg on the inactive system.
varyonvg <diskgroup> : on active node to put the reserve back in place

Can be done without any outage to the diskgroup.

By the way, how do you normally sync up your volume groups in your cluster when you add a disk or filesystem?
#AIX-Forum