Unplanned switch - overview - assure_mimix - 10.0

Assure MIMIX Operations Guide

Product type
Software
Portfolio
Integrate
Product family
Assure
Product
Assure MIMIX™ Software
Version
10.0
Language
English
Product name
Assure MIMIX
Title
Assure MIMIX Operations Guide
Copyright
2023
First publish date
1999
Last updated
2024-03-12
Published on
2024-03-12T11:06:36.794496

In an unplanned switch, the source system is assumed to be unavailable. An unplanned switch is generally required when the source system fails and, in order to continue normal operations, you must switch users to a backup system. (Typically MIMIX is configured so that the target for replication is your backup system.)

You must run an unplanned switch from the target system. MIMIX performs a controlled shutdown of replication processes on the target system. The controlled shutdown allows all apply processing to catch up before the apply processes are ended.

In an unplanned switch of a data group that uses remote journaling, the default behavior is to end the RJ link.

Following an unplanned switch: Once the failed source system is recovered, the following actions should be completed:

  • You should perform journal analysis on that system before restarting the data group or user applications. Journal analysis helps identify any possible loss of data that may have occurred when the source system failed. Journal analysis relies on status information on the source system about the last entry that was applied. This information will be cleared when the data group is restarted.

  • Communication between the systems must be active before you restart the data group. The switch process is complete when you restart the data group.   When the data group is restarted, MIMIX notifies the source system that it is now the temporary target system.

  • New transactions are created on the temporary source system (the backup system) while the production system (the temporary target system) is unavailable for replication. After you have completed journal analysis, you can send these new transactions to the production system to synchronize the databases. Once the databases are synchronized, you must run the switch process again to revert to the normal roles before allowing users onto the production system.

When the data group is started after a switch, any pending transactions are cleared. The journal receiver is already changed by the switch process and the new journal receiver and first sequence number are used.