11-03-2015 07:31 PM
In Veritas Cluster Server (VCS), an agent monitors the status of resources. A resource can be in one of the following states:
However, sometimes a problem can occur. If the agent cannot determine the status of the resource within a predetermined time, the agent restarts the resource.
Problems can occur, when:
• If an ONLINE resource is reported OFFLINE without any offline operation performed from within VCS, or the resource is offlined from outside of VCS control, also known as unexpected Offline.
• If FaultOnMonitorTimeouts (FOMT) is reached i.e., the monitor entry point has timed out FOMT times.
• If the OFFLINE entry point is ineffective and the monitor entry point timed out after that. This is not the case of unexpected Offline.
The error message V-16-2-13074 shows in the log if all the following conditions are met in order:
• Monitored resource is faulted.
• ManageFaults != NONE.
• Clean entry point is executed successfully. (If clean fails then clean entry point is called only until the CleanRetryLimit attribute is reached).
• It is not the case of unexpected Offline.
• The agent restarts the resource. (It happens only if the RestartLimit attribute is not reached)
This error applies to all platforms and releases.
To solve this issue:
• Check whether the return status of the monitor returns OFFLINE.
• Check whether the monitor is timed out. This is because the ineffective Offline or the system heavily loaded results in multiple monitor timeouts.
For details about monitoring the resources, see:
How VCS handles resource faults Controlling VCS behavior
For more information, go to the SORT website.