RH cluster and NetWorker

Not yet, but we will be starting our test processes on RHEL5 HA clustered NetWorker servers within the next month or so.

Thus we would also be very interested in other people experiences.

Felix

Hi Felix,

I just got the boxes and waiting for cluster license from RH.  Then the fun begines

Hi,

We have recently requested EMC to officialy support RedHat High Avilability cluster as we would like to use this cluster solution in our new NetWorker infrastructure.

I see you post is from last year, did you had any answer from EMC on your side ?

Did you succeed to install NW on RH cluster ?

Kind regards,

Fabian

I did, but we gave up that idea at the end. RH cluster does not seem to be very good product and that explains why it is not being seen so much deployed elsewhere (here I refer to backup server being in clsuster and not client).

Hi,

we have 2 RH clusters with NW 7.6.1 and 7.6.3 in production, but we have started to have some problems with stability of cluster after migration of more backups to them.Problem is that function nsr_monitor in nw_vcs script is checking response of nsrd every minute and sometimes happened that nsrd did not responed. For this reason exit code 100 is send to veritas cluster software, which will affect NW. All running groups are terminated and sometimes happened that cluster service switched to another node.

This is really not good, so I have changed nsr_monitor function that only echo "exit 100" is written to our debug log and function sends exit code 110 to veritas custer software. It prevents temporary problems with responding of nsrd.

Cheers,

Edo

What you have is VCS, and not red hat cluster (Red Hat comes with its own cluster solution).  Values for rpcinfo timeout are set on normal usage or not so heavy load.  I also had the same issue with HPUX cluster and had to modify default values in legato.monitor script.