Adding disks on HACMP -How not to get warning - Aix

This is a discussion on Adding disks on HACMP -How not to get warning - Aix ; When I add disk in 2 node cluster I Sync with varyonvg -b -u and importvg -L I always get following warning added in hacmp veriifcation why? WARNING: The HACMP timestamp file for shared volume group: apsvolumeg is inconsistent with ...

+ Reply to Thread
Results 1 to 2 of 2

Thread: Adding disks on HACMP -How not to get warning

  1. Adding disks on HACMP -How not to get warning

    When I add disk in 2 node cluster
    I Sync with varyonvg -b -u and importvg -L
    I always get following warning added in hacmp veriifcation why?
    WARNING: The HACMP timestamp file for shared volume group: apsvolumeg
    is inconsistent
    with the time stamp in the VGDA for the following nodes: qscf2007A
    qscf2007B

    How to avoid this at first place?
    I don't need how to fix this.

    Thank you very much in prior anticipation.


  2. Re: Adding disks on HACMP -How not to get warning

    On Oct 31, 10:13 am, kakir...@hotmail.com wrote:
    > When I add disk in 2 node cluster
    > I Sync with varyonvg -b -u and importvg -L
    > I always get following warning added in hacmp veriifcation why?
    > WARNING: The HACMP timestamp file for shared volume group: apsvolumeg
    > is inconsistent
    > with the time stamp in the VGDA for the following nodes: qscf2007A
    > qscf2007B
    >
    > How to avoid this at first place?
    > I don't need how to fix this.
    >
    > Thank you very much in prior anticipation.


    This has happened to me anytime I update a VG without using the
    cluster methods...ie: growing a filesystem using chfs at the command
    line on the server accessing the VG. If that is done then the VGDA
    records on the VG are out of sync with what HACMP thinks it should
    be. The same thing might happen when you add a disk...so you should
    try using the smitty cspoc => HACMP Physical Volume Management
    instead.

    I used to correct the timestamp error by taking the cluster and
    running extended verification with fix problems enabled. Later I
    found I can do the same with cluster running by just slightly growing
    a shared filesystem on the VG using smitty hacmp=> CSPOC => HACMP
    Logical Volume Management => Shared Filesystems. HACMP will then
    update itself with the new VGDA timestamp after it does it's thing.
    Maybe there's another way but this was quick and easy the few times I
    needed it.

    cheers


+ Reply to Thread