Return values for entry points
Prev Agent entry point overview Next

Return values for entry points

The following table summarizes the return values for each entry point.

Table: Return values for entry points

Entry Point

Return Values

Monitor

C++ Based Returns ResStateValues:

  • VCSAgResOnline

  • VCSAgResOffline

  • VCSAgResUnknown

  • VCSAgResIntentionalOffline

Script-Based Exit values:

  • 99 - Unknown

  • 100 - Offline

  • 101-110 - Online

  • 200 - Intentional Offline

  • Other values - Unknown.

Info

0 if successful; non-zero value if not successful

Online

Integer specifying number of seconds to wait before monitor can check the state of the resource; typically 0, that is, check resource state immediately.

Offline

Integer specifying number of seconds to wait before monitor can check the state of the resource; typically 0, that is, check resource state immediately.

Clean

0 if successful; non-zero value if not successful

If clean fails, the resource remains in a transition state awaiting the next periodic monitor. After the periodic monitor, clean is attempted again. The sequence of clean attempt followed by monitoring continues until clean succeeds or CleanRetryLimit is not reached if it is set to non-zero value.

For detailed descriptions of internal transition states, See State transitions.

Action

0 if successful; non-zero value if not successful

Attr_changed

None

Open

None

Close

None

Shutdown

None

imf_init

0 if successful; 1 if unsuccessful

imf_register

0 if successful; 1 if unsuccessful

imf_getnotification

0 if successful; 1 if failure; 3 if interrupted (failure case); 4 if critical failure

migrate

  • An integer in the range of 0 to 100. The typical return value is 0. If the return value is not zero, the agent framework waits for the number of seconds indicated by the (return value * 10) to call the monitor entry point for the resource. For example, for a return value of 1, agent framework schedules monitor after 1*10=10 seconds. Similarly, for a return value of 5 monitor is scheduled after 50 seconds.

  • 255 indicating that migration verification has failed and there is no need to schedule a monitor to verify whether resource has migrated. The subsequent monitor is a scheduled based on the MonitorInterval value.

All other values in the range of 101 to 254 are reserved for future use. Agent framework ignores any value returned between this range and returns to previous state to continue with rest of the operations. Refer to MigrateWaitLimit and MigrateTimeout, before implementing this entry point.

See MigrateTimeout.

See MigrateWaitLimit.

See SupportedOperations.

meter

  • 0 - Indicates that meter entry point has completed successfully.

  • 255 - Indicates that meter entry point has failed.

    METER FAILED flags will be set if meter entry point fail or timeout.

  • 254 - If the meter entry point fails with this return value, then it is treated as critical fault and metering is disabled for the resource until the agent restarts

All other values in the range of 1 to 253 are reserved for future use. The Agent framework considers those values as failure with unsupported value and sets the METER FAILED flag. These failure with unsupported value will not be counted against the MeterRetryLimit, so meter entry point should not use these values.

See MeterRetryLimit.
Prev Up Next
About the meter entry point Home Considerations for using C++ or script entry points