How to Maintain the HADB Database

When you want to run HADB maintenance commands that initiate a rolling restart of the HADB nodes, the HADB resource probe in the fault monitor must be disabled before the maintenance commands are executed and then enabled after the commands and the rolling restart have completed.

Steps

1. Disable the fault monitor.

   # scswitch -n -M -j resource

2. Run the commands that might initiate a rolling restart.

   The hadbm subcommands that might result in a rolling restart include set, restart, and addnodes.

3. Reenable the fault monitor.

   # scswitch -e -M -j resource

Alternative Method of Maintaining the Database

Alternatively, the HADB resource can be disabled and HADB started outside of Sun Cluster control while maintenance commands are being performed.