Took mksysb's from a working cluster with 2 nodes (HACMP was down at the time of the backup).
I restored these to another cluster, changed the hostname and IP addresses, and changed HACMP adapters to the correct addresses.
When I try to start HACMP, it never starts.
I found the following error in /tmp/clstrmgr.debug
Wed Jan 29 09:48:17 RdInit: Initializing group services interface
Wed Jan 29 09:48:17 pgs_init failed, rc=19
Wed Jan 29 09:48:22 RdInit: Initializing group services interface
Wed Jan 29 09:48:22 pgs_init failed, rc=19
This repeats over and over....every few seconds until I force a shutdown of the cluster.
I've tried removing the cluster configuration and redoing manually again, and still get the same problem.
I did find the following on the web:
Certain high somaxconn settings (in the range of 32768-65535) will cause grpsvcs daemon to be blocked, with no daemons able to form groups with it. The symptom this was discovered under was clstrmgr failing to start fully, with the following errors looping in clstrmgr.debug every five seconds:
RdInit: Initializing group services interface
pgs_init failed, rc=19
Checking lssrc -ls grpscvs showed the daemon active but no
groups formed. Forcing the cluster back down and setting the network setting to a lower value (such as no -o somaxconn=2048) allowed grpsvcs to function and the cluster to come up normally
I reset the somaxconn to 2048 as mentioned, but it doesn't solve the problem....and since this cluster works fine on the other boxes, I don't think that this is the problem.
Any help would be appreciated!
Thanks
I restored these to another cluster, changed the hostname and IP addresses, and changed HACMP adapters to the correct addresses.
When I try to start HACMP, it never starts.
I found the following error in /tmp/clstrmgr.debug
Wed Jan 29 09:48:17 RdInit: Initializing group services interface
Wed Jan 29 09:48:17 pgs_init failed, rc=19
Wed Jan 29 09:48:22 RdInit: Initializing group services interface
Wed Jan 29 09:48:22 pgs_init failed, rc=19
This repeats over and over....every few seconds until I force a shutdown of the cluster.
I've tried removing the cluster configuration and redoing manually again, and still get the same problem.
I did find the following on the web:
Certain high somaxconn settings (in the range of 32768-65535) will cause grpsvcs daemon to be blocked, with no daemons able to form groups with it. The symptom this was discovered under was clstrmgr failing to start fully, with the following errors looping in clstrmgr.debug every five seconds:
RdInit: Initializing group services interface
pgs_init failed, rc=19
Checking lssrc -ls grpscvs showed the daemon active but no
groups formed. Forcing the cluster back down and setting the network setting to a lower value (such as no -o somaxconn=2048) allowed grpsvcs to function and the cluster to come up normally
I reset the somaxconn to 2048 as mentioned, but it doesn't solve the problem....and since this cluster works fine on the other boxes, I don't think that this is the problem.
Any help would be appreciated!
Thanks