Detecting Application Failures
Cluster members use IP sockets to detect failures—when a socket suddenly closes during data transmission, the server is marked as “failed” and all the services offered by the server are removed from the JNDI tree. When servers aren't actively communicating via open sockets, WebLogic Server detects failures by monitoring the server heartbeats, which by default are sent every ten seconds by every cluster member. Cluster members use the IP unicast (or multicast) to broadcast and revive the heartbeat messages. Once a server misses three heartbeats from a peer server, it marks the status of the other server as failed and updates its local JNDI tree by removing the failed server's services.
In the following sections, ...
Get Oracle WebLogic Server 11g Administration Handbook now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
