Hi Tony, Thanks for the response. The required packages are "SUNWscucm SUNWudlm SUNWudlmr SUNWscmd". These are installed on the system. /var/run/nodelist is not available on *both* the nodes after restart but the metaset output is not same on the two nodes. 1. On host1: Multi-owner is NOT recognised, Master is recognised, host2 is not a member root at host1 # metaset -s oradg Multi-owner Set name = oradg, Set number = 1, Master = host1 Host Owner Member host1 Yes host2 No
2. On host2: Multi-owner is recognised, Master is NOT recognised, host2 is not a member root at host2 # metaset -s oradg Multi-owner Set name = oradg, Set number = 1, Master = Master and owner information unavailable until joined (metaset -j) Host Owner Member host1 multi-owner Yes host2 No Now /var/run/nodelist needs to be created on both the nodes and then host1 is to be joined. root at host1 # metaset -s oradg -M -j After this, host2 gets recognised as a member (-M option is required for this to happen). root at host2 # metaset -s oradg Multi-owner Set name = oradg, Set number = 1, Master = host1 Master and owner information unavailable until joined (metaset -j) Host Owner Member host1 multi-owner Yes host2 Yes Now, host2 can be joined using "metaset -s oradg -M -j". After this the output is as below on both the nodes. Host Owner Member host1 multi-owner Yes host2 multi-owner Yes The cluster itself has been setup using the EIS procedure. Even with the above manual procedure there is a availability problem. If the system is rebooted and host1 does not comeup due to any reason, then host2 cannot provide service since it is not a member/owner. After the reboot I checked the /var/adm/messages for any hints of problems. The following entries were the only ones which seemed connected in any manner. /var/adm/messages on host1 immediately after restart ====================================================== Aug 18 10:45:47 host1 Cluster.CCR: [ID 988885 daemon.error] libpnm error: can't connect to PNMd on host2. /var/adm/messages on host2 immediately after restart ====================================================== Aug 18 10:46:08 host2 svc.startd[8]: [ID 652011 daemon.warning] svc:/system/cluster/scsymon-srv:default: Method "/usr/cluster/lib/svc/method/svc_scsymon_srv start" failed with exit status 96. Aug 18 10:46:08 host2 svc.startd[8]: [ID 748625 daemon.error] system/cluster/scsymon-srv:default misconfigured I came across a related post at below link which hasn't reached any logical conclusion http://forum.sun.com/jive/thread.jspa?threadID=76461&messageID=278172 regards Shiv This message posted from opensolaris.org
