The Rulebase server that responds to requests from clients is the primary server. The other Rulebase servers are secondary servers. The secondary servers poll the

IDS_RB_GROUPS

table periodically to verify the status of the primary server. If the primary server goes down, one of the secondary servers becomes the primary server.

You can use the server groups in a distributed database environment where multiple database servers run on different network nodes. All the database servers connect to a shared disk sub-system using a storage area network and appear as a single unified database. The database remains available even if one node is operational. You can start or stop additional nodes transparently without affecting the connectivity or data integrity.

Before you set up high availability using the server groups, you must consider the following aspects of the Rulebase servers:

The Rulebase server must access a robust database instance. The database instance must remain available even if there is one active node in the network.
Only one Rulebase server responds to the Rulebase requests even though the server groups might contain multiple Rulebase servers.
You can shut down all the servers in a server group using the
idsdown
script. You cannot shut down an individual secondary Rulebase server. You can only kill an individual secondary Rulebase server.
You must always start the Search and Rulebase servers as separate servers and establish communication between them through sockets. Don't start the Search and Rulebase servers together.
You must restart the Rulebase servers only with the
-o
switch. You must ensure that the Rulebase servers don't start automatically by the Console server.
To assign Rulebase servers to the server groups, use the
-g
switch.