An asymmetric configuration consists of two Adaptive Servers running on separate machines. The primary Adaptive Server performs the work during day-to-day operations, while the secondary Adaptive Server is prepared to take over the work during a system failure or a scheduled maintenance. The secondary companion is an independent Adaptive Server, and can have its own applications running. To configure for failover, the secondary companion must be a newly installed Adaptive Server, and cannot yet have any user logins or user databases. After configuration is complete, you can add user logins and databases to the secondary companion.
When you install and configure Adaptive Server for failover, Adaptive Server is in single-server mode. Use sp_companion to change it from single-server mode to a companion server in an asymmetric setup. See the Adaptive Server Reference Manual for information about sp_companion.
The primary companion issues messages similar to the following when you configure Adaptive Server for failover:
sp_companion "MONEY1", configure Step: Server 'PERSONEL1' is alive and cluster aware Changing physical name of server 'SYB_HACMP' from 'PERSONEL1' to 'MONEY1' Step: Access verified from Server:'PERSONEL1' to Server:'MONEY1' Step: Server 'MONEY1' is alive and cluster aware Changing physical name of server 'SYB_HACMP' from 'MONEY1' to 'PERSONEL1' Step: Access verified from Server:'MONEY1' to Server:'PERSONEL1' Step: Companion servers configuration check succeeded Step: Server handshake succeeded Step: Master device accessible from companion Step: Added the servers 'PERSONEL1' and 'MONEY1' for cluster config Step: Server configuration initialization succeeded Step: Synchronizing server logins from companion server Step: Synchronizing remoteserver from companion server Step: Synchronizing roles from companion server Step: Synchronizing server-wide privs from companion server Step: User information syncup succeeded Step: Server configured in normal companion mode
Figure 3-1 describes an asymmetric configuration:
Figure 3-1: Asymmetric configuration in a high availability system
In this setup, MONEY1 is the primary companion and fails over to PERSONEL1, the secondary companion. Both disks are visible to machine HUM1, which connects to machine FIN1 with a dual-ported SCSI. Because this is an asymmetric setup, PERSONEL1 cannot fail over to MONEY1. Disk 1 must be a shared disk, and disk 2 can be a local disk.
See the configuration chapter for your platform for information about configuring Adaptive Server for an asymmetric setup.
