Backend server or backend node fails

Universal Cache User Guide : Failure handling in Universal Cache : Backend server or backend node fails

If the backend server or the backend node fails, replication on the subscriptions ends also. To recover, proceed as follows:

Procedure

1 Restart the backend server and recover the database.

2 Restart the IBM InfoSphere CDC instance.

3 Restart mirroring (replication) on the subscriptions.

Note It may be possible to automate the above steps using the backend specific tools and procedures.

Results

When restarted, the replication resumes and the databases are resynchronized. Replication continues normally.

If this failure happens, the solidDB^® server continues to process transactions until it reaches the limit specified by the LogReader.MaxLogSize parameter.