High Availability Guide : Watchdog sample : Failure situations and Watchdog actions : Primary is down
  
Primary is down
Scenario
All connections to the Primary server are broken.
Remedy
When the Primary is down, switch the Secondary to be the new Primary and set the new Primary to the PRIMARY ALONE state. Later, the old Primary can become a new Secondary.
1 Watchdog instructs Server2: HSB SET PRIMARY ALONE Applications switch from Server1 to Server 2. HSB SET STANDALONE
2 After Server1 is fixed, Server 1 is brought back up as Secondary. Watchdog instructs Server 2:
HSB NETCOPY
HSB CONNECT
If the transaction log fills up, you may have to switch the PRIMARY ALONE server to STANDALONE. In that case, you will also need to execute HSB NETCOPY before you reconnect the servers. If the transaction log does not fill up, then you must skip the NETCOPY command.
Symptoms
Applications cannot connect to the Primary. Also, the watchdog poll fails at the Primary. The HSB state of the secondary server is SECONDARY ALONE.
How to recover when the primary is down
About this task
To allow the “HotStandby” (Secondary server) to replace the Primary, do the following:
Procedure
1 Set the new Primary server to PRIMARY ALONE state by using the command:
ADMIN COMMAND 'hotstandby set primary alone';
2 Reconnect applications to the new Primary.
3 Start using applications.
4 Fix and start the old Primary server as new Secondary server.
5 If necessary, copy the database from the new Primary to the new Secondary using command:
ADMIN COMMAND 'hotstandby netcopy';
Read Synchronizing primary and secondary servers for details.
6 Reconnect the new Primary to the new Secondary using the command:
ADMIN COMMAND 'hotstandby connect';
See also
Failure situations and Watchdog actions