0x10524 Fault-Tolerant Database Synchronization Failure

Document ID : KB000049935
Last Modified Date : 14/02/2018
Show Technical Document Details

Description:

Spectrum's database backup and synchronization feature has a default timeout of 60 minutes. If the total time needed to backup and synchronize the database takes longer than the value specified in the OLBTimeoutInMinutes (0x1295c) attribute the process will fail and event/alarm 0x10524 will be generated. You may notice that a scheduled or manual online backup starts, pauses, and resumes normal operation, but then stays in synchronization for an hour before issuing the 0x10524 Fault Tolerant Database Synchronization Failure event and alarm. This happens when the SpectroSERVER database is quite large.

Solution:

There is an attribute on the VNM model named OLBTimeoutInMinutes (0x1295c). This is set to 60 minutes by default, which is the overall timeout value for the backup to complete and for the database synchronization to occur. This value can be changed based on the total length of time for this process. As the synchronization is the last piece of the backup, the RCPD.OUT file in the $SPECROOT/SS/ directory can be reviewed to see how long this process takes. Based on the length of this process observed in RCPD.OUT file, you can judge what the value should be for the OLBTimeoutInMinutes (0x1295c) attribute. The attribute can be changed through CLI or the Attribute Editor. This will resolve the 0x10524 Fault Tolerant Database Synchronization Failure event and alarm when it is due to the backup process taking a long time to complete.

Note: Administrative notes should be made as to why this attribute was changed, how it was done and what value it was changed to. Future upgrades, service pack installs or patch installs may revert back to the default value.

The current probable cause statement on the 0x10524 alarm does not list the above problem and solution description, but will be updated to include it.