Restarting HDR or RS clusters if the secondary server fails

If you must restart HDR or an RS cluster after a failure of a secondary server, in addition to starting the secondary server, you might need to perform a logical log restore on the secondary server.

The steps assume that you have been backing up logical-log files on the primary database server as necessary since the failure of the secondary database server.

Table 1. Steps in restarting after a failure on the secondary database server
Step On the primary On the secondary
1. The primary database server must be in online mode. oninit

If you receive the following message in the message log, continue with step 2:

DR: Start Failure recovery from tape
2. command

ON-Bar command

onbar -r -l