Last Modified: July 06, 2012 - Raj Patel ( rajpat@us.ibm.com ) Recovering caavg_private when repository disk no longer reachable. Possible Cause: 1) Power Outage. 2) If it fails to start with below error: Command: $ clstartstop -start -n [cluster-name] -m [node] Error: ====== cluster_utils.c does_cluster_exist 5923 return = -1, Failed to find cluster, rc = -1, errno=109, code=5907 This node is not part of specified cluster annecluster01. 3) Check ODM entry for name = "cluster0" # odmget -q name=cluster0 CuAt "attribute" & "value" should not be blank. CuAt: name = "cluster0" attribute = "clvdisk" value = "99681e7b-9e56-0a35-be23-62c6677d9d90" type = "R" generic = "DU" rep = "s" nls_index = 2 If it is blank, use below to re-populate within oem_Setup_env. # "clusterconf -r hdisk#" This should populate the ODM with clvdisk attribute and be able to then restart the SSP Cluster. 4) Possible solution: $ varyonvg caavg_private $ startsrc -s pool $ startsrc -s vio_daemon $ stopsrc -g caa $ startsrc -g caa 5) Rerun below: clstartstop -start -n [cluster-name] -m [node] 6) get status $ cluster -status -clustername [cluster-name] 7) wait for 10 minutes and get status again. $ cluster -status -clustername [cluster-name]