[Veritas-ha] VCS 1.3.0 with VVR---Long failover time solaris 2.6

[Sampath K Kari]

Hi all,

configuration:
we have a 2 node asymmetric VCS 1.3.0 cluster. Both nodes run solaris
2.6, VM 3.1.1, VXFS 3.3.3. we replicate from this cluster using Veritas
Volume Replicator (VVR) to another 2 node cluster of same configuration.
There are 60 RVolume type resources. DiskGroup type resources have the
attributes "StartVolumes" and "StopVolumes" set to 0.

Problem:

We had a crash on 1 node in the cluster causing a VCS failover to other
node.  Total failover time was 34 minutes.  It took 31 minutes for all
the 60 RVolume resources to come online.

Is there any  way we can cut short this failover time by improving the
time to bring the 60 RVolume type resources?

Thanks a lot for any pointers.

-Sampath


-------------- next part --------------
A non-text attachment was scrubbed...
Name: sampath_k_kari.vcf
Type: text/x-vcard
Size: 241 bytes
Desc: Card for Sampath Kari
Url : http://mailman.eng.auburn.edu/pipermail/veritas-ha/attachments/20011017/75b4c6ec/sampath_k_kari.vcf