Understanding the Sun Cluster HA for Oracle E-Business Suite Fault Monitor

This section describes the Sun Cluster HA for Oracle E-Business Suite fault monitor's probing algorithm or functionality.

For conceptual information on fault monitors, see the Sun Cluster Concepts Guide.

Resource Properties

The Sun Cluster HA for Oracle E-Business Suite fault monitor uses the same resource properties as resource type SUNW.gds. Refer to the SUNW.gds(5) man page for a complete list of resource properties used.

Probing Algorithm and Functionality

• Concurrent Manager Server

  • Sleeps for Thorough_probe_interval

  • Test whether at least one FND (Concurrent Manager) process is running for CON_APPSUSER. If this fails, then the probe will restart the Concurrent Manager Server resource.

  • Test whether the probe can still connect to the Oracle Database. If this fails, then the probe will restart the Concurrent Manager Server resource.

  • Calculate the number of concurrent processes running as a percentage of the maximum number of concurrent processes allowed, and test whether that percentage is less than CON_LIMIT, when the Concurrent Manager Server resource was defined. If the percentage is less than CON_LIMIT, then the probe will restart the Concurrent Manager Server resource.

  • If all Concurrent Manager Server processes have died, pmf will interrupt the probe to immediately restart the Concurrent Manager Server resource.

  • If the Concurrent Manager Server resource is repeatedly restarted and subsequently exhausts the Retry_count within the Retry_interval, then a failover is initiated for the Resource Group onto another node.

• Forms Server

  • Sleeps for Thorough_probe_interval

  • Test whether the f60srvm process is running for FRM_APPSUSER. If f60srvm is found, then test whether f60webmx process is running. If f60webmx is not found, then the probe will retest after another iteration of the probe to determine whether f60webmx is still missing, because f60srvm usually restarts f60webmx. If after two successive probes, f60webmx is still missing or f60srvm is not found on any probe, then the probe will restart the Forms Server resource.

  • If the Forms Server resource is repeatedly restarted and subsequently exhausts the Retry_count within the Retry_interval then a failover is initiated for the Resource Group onto another node.

• Report Server

  • Sleeps for Thorough_probe_interval

  • Test whether rwmts60 process is running for REP_APPSUSER. If this fails, then the probe will restart the Forms Server resource.

  • If the Report Server resource is repeatedly restarted and subsequently exhausts the Retry_count within the Retry_interval then a failover is initiated for the Resource Group onto another node.