Sybase’s failover includes the following logs. These logs are helpful for investigating and diagnosing errors encountered during HACMP failover:
/tmp/hacmp.out – contains output of the HACMP activities, as well as the output from the RUNHA_server_name.sh monitoring script. For general HACMP failure, search for the string “ERROR”. For output of the RUNHA_server_name.sh script, search for “SYBASE HA MONITOR”.
After determining the reason for the failure, correct it, then go to the Cluster Recovery Aids screen of SMIT and do a Recover From Script Failure, before continuing.
If a node does not include a sufficient amount of space in a particular file system, HACMP hangs in the middle of a failover or failback process, which results in a config_too_long lock. If this occurs, you must clean up the full directories, then start SMIT and move to the Cluster Recovery Aids screen and perform a Recover From Script Failure before continuing.
$PRIM_CONSOLE_LOG – the location of this log is defined in the RUNHA_server_name.sh monitoring script. This error log includes the Adaptive Server information from the last execution of the RUNHA_server_name.sh script.
