FailOver and FailBack are the significant recovery processes of virtual machine replication which are actually involved in the Backup and Disaster Recovery process. If you are curious about FailOver & FailBack and want to know “Why do we need FailOver and FailBack” go through my blog “FailOver & FailBack – A Significant Process During Disaster”.
The process FailOver & FailBack are enriched by their sophisticated and easy to use options which increases the usability not only for disaster recovery but for testing and maintenance purposes also.
Let us discuss the actual mechanism of FailOver & FailBack and its options that are combined together to make the process more powerful.
FAILOVER
Fail Over is a process of switching the replica machine on, while the source machine is affected because of a disaster or the source is in maintenance mode. During the failover the replica machine will act as source machine and handle all the processes.
If the replica job configured with Network and IP-Remapping then the replica will up and run with that Network & IP configurations during the FailOver state, If not you have to manually assign Network and IP for the replica.
It is possible to FailOver the replica to any of its available recovery point. It is also possible to FailOver group of replicas which are actually dependant on one another that are available in the configuration.
The data added during the FailOver state will be maintained in a snapshot delta file which will be committed later based on the need.
FailOver is a temporary state that may further be finalized by any of the below processes:
1→ Undo FailOver
2→ Permanent FailOver
3→ FailBack
Undo FailOver
Undo FailOver is a process of reverting back the replica to its original state before FailOver. The data added during the FailOver state will be flushed after the Undo FailOver.
Undo FailOver might be useful if the FailOver was initiated for testing purpose or if you got the original production server back after the troubleshoot process and you don’t really want the data that has been added during the FailOver state.
Permanent FailOver
Permanent FailOver is a process of making the replica machine as the source machine permanently. In case if the source machine is not recoverable after a disaster then you can go for permanent FailOver option.
All the changes made during the failover state will be committed to the disk once the permanent failover done. Further incremental will be blocked for the replica after the permanent failover and the VM will be excluded from the replica job.
FailBack
FailBack is a process of restoring the source machine in the production environment from its replica. If the malfunctioned original machine recovered back after the troubleshoot process and you want to resume the operation using the same with the latest data then the FailBack process will be helpful on the aforementioned scenario.
There are numerous ways of FailBack:
We can FailBack to the same production machine in which only the difference between the two disks will be committed.
We can FailBack to the same host as a new machine in which replica or original machine available.
We can FailBack to other host as a new machine with the latest data and ready to use as production machine.
Once after the FailBack process is done the production machine will up and run with the latest changes and will handle all the I/O of production operations.
The FailBack state may further be finalized by any of the below processes:
1→ Undo FailBack
2→ Commit FailBack
Undo FailBack
Undo FailBack is a process of reverting back a replica to the FailOver state. The current FailBack snapshot will be deleted and reverted the replica to FailOver state.
Undo FailBack is useful in the below scenarios:
If the FailBack process is not successful and need to secure the data added during the FailOver state then Undo FailBack could be done.
If the FailBack process is successful and the Failed Back machine is not working as expected then the Undo FailBack could be done.
Commit FailBack
Commit FailBack is the confirmation process of FailBack. If the Failed Back VM is recovered properly and running as expected then we can confirm the FailBack by commit FailBack.
When we perform Commit FailBack, that means we are confirming to resume the Failed Back VM as original production VM which will gets back to normal production mode.
Once Commit FailBack done the original VM will be excluded from the list of replica jobs.
To get experience in Vembu VMBackup, please give a try:
https://www.bdrsuite.com/vembu-bdr-suite-download/
Got questions? Email us at: vembu-support@vembu.com for answers.
Follow our Twitter and Facebook feeds for new releases, updates, insightful posts and more.
