For up-to-date product documentation, see the IBM MobileFirst Foundation Developer Center.

Lifecycle of a server farm node

We can configure heartbeat rate and timeout values to indicate possible server problems among farm members by triggering a change in status of an affected node.

Registration and monitoring servers as farm nodes

When a server configured as a farm node is started, the administration service on that server automatically registers it as a new farm member.

When a farm member is shut down, it automatically unregisters from the farm.

A heartbeat mechanism exists to keep track of farm members that might become unresponsive, for example, because of a power outage or a server failure. In this heartbeat mechanism, MobileFirst runtimes periodically send a heartbeat to MobileFirst administration services at a specified rate. If the MobileFirst administration service registers that too long a time has elapsed since a farm member sent a heartbeat, the farm member is considered to be down.

Farm members that are considered to be down do not serve any more requests to mobile applications.

Having one or more nodes down does not prevent the other farm members from correctly serving requests to mobile applications nor from accepting new management operations that are triggered through the MobileFirst Operations Console.

Configuring the heartbeat rate and timeout values

We can configure the heartbeat rate and timeout values by defining the following JNDI properties:

For more information about JNDI properties, see List of JNDI properties for MobileFirst Server administration service.

Parent topic: Installing a server farm