If a reference job is performed on the standby server and the primary server is switched, this may impact the main job from the point of view of load and conflict. This is because, on the new primary server (that is, the original standby server), both the main job that was being executed on the original primary server and the reference job that was being continued on the original standby server will be processed.
Therefore, to degrade the reference job (so that the impact on the main job is reduced), you can select the user exit below to disconnect the reference job that was performed on the original standby server.
Post-switch command
Note
If continuing with the referencing job after switching the primary server, give careful consideration to the server resource estimates, and the likely impact on performance.
If an issue such as a log transfer network failure obstructs the continuation of a job on the primary server, the standby server may be automatically disconnected from the cluster system.
Therefore, if operating the reference job on the assumption that the connection destination is the synchronous standby server, you can select to temporarily stop the job by using the user exit or the feature below, so that unexpected referencing of past data does not occur as a result of the disconnection.
Pre-detach command
Forced stoppage of the standby server instance on disconnection (specify in the parameter of the server configuration file)
Additionally, if the standby server is incorporated into the cluster system, reference jobs can be started or resumed by using the user exit below.
Post-attach command
See
Refer to "2.6 Creating a User Exit for a Database Server" or "Appendix C User Commands" for information on each user exit.
Refer to "A.4.1 Server Configuration File for the Database Servers" for information on the server configuration file of the database server.
Point
Mirroring Controller will continue processing regardless of the processing result of the above user exits and features.
