Skip to main content

CommServe Recovery Using Disaster Recovery (DR) Backups (NOT FAILOVER)

 

11.28.72

OBS: I would like to test the DR Backup and run the procedure at least once a year. 

 

I'm following the instructions here to test the DR backup in case of commserve disaster recovery.

In the event of a disaster, when the production CommServe host becomes unusable or inaccessible, or if the CommServe host is moved to a new hardware, the latest Disaster Recovery (DR) backups can be used to rebuild the CommServe host.

 

At step 4, we should use the CommServe Recovery Assistant Tool to restore the most recent Disaster Recovery (DR) backup from the production CommServe database to a standby or new CommServe host to resume the CommCell operations. For more information on recovering the CommServe database, see Recovering the CommServe Database Using the Recovery Assistant Tool.

 

At step 2 from Recovering the CommServe Database Using the Recovery Assistant Tool for Windows page (here) we got the following instruction: 

 

“Stop services on the original production CommServe host, if it is active. For more information on stopping services, see Stopping a Service.”

 

The original production CommServe host is active, but I would like to do the procedure without stoping the services on the production CommServe host. Is it possible? Is there any risk that?  

 

The “standby” server uses a different IP and Host Name for the CommServe host. They are vmware VM and They are not on the same network, but they are accessible from each other.

Hi @Eduardo Braga ,

 

I would suggest to isolate the standby Server Server from the Network or bring the Standby VM out of the domain to test the DR.


Hi Eduardo,

 

As Nutan has already advised, as long as the Test Standby CS is isolated within its own network with no communication whatsoever to the production environment you should be fine leaving the Active/Production CS services running.

 

The recommendation is placed in documentation to ensure no clashing or potential split-brain events happen when another set of CS Services with the same setup as the production CS becomes active in the network, as you know, only 1 Machine hosting the CS roles is allowed at any time.


Reply