cancel
Showing results for 
Search instead for 
Did you mean: 

Backup Exec Replication & Recovery

akuma22
Level 3

Hi,

Using BE 2012 SP1, I've successfully setup a job to backup and replicate the de-duped data to our DR site.

The master server is configured as a Central Administration Server (CAS) and the DR site server is setup as a Managed Media Server (MMS).

Backup, replication and restores are working fine from both sites but I noticed that when the CAS server becomes unavaialble, the MMS is no longer able to perform any backup/restore operations and some of the BE services are stopped.

I would like to be able to recover backups from the DR site in the event that the master site/server is offline. What must I configure to ensure the MMS server remains operational even if the CAS server is offline?

Thanks

1 ACCEPTED SOLUTION

Accepted Solutions

Colin_Weaver
Moderator
Moderator
Employee Accredited Certified

The Opt Dup configuration gives two choices.

 

1) Original backup done to MMS/MBE, OptDup to CAS. In which case if CAS (or link to CAS goes down your initial backups may not run as the MBE depends to a certain extent on the CAS database

2) Original Backup done to CAS, OptDeup to MMS/MBE. In this scenario your initial backups should always run, the duplicates will fail if the WAN link or MMS/MBE is unavailable. Again the CAS going unavilable would also kill your backups,

 

In either scenario you could spend time if a failure happens, doing a reconfigure of whichever server is still running back to stand alone mode, which could then be used for Backups or Restores (assuming your backed up data is available. However you would probably be better off spending time fixing the problem and then just re-run the jobs once fixed.

 

Basically this is not a multi-master model so cannot achive full resilience and some manual labour will always be needed to recover from a lost media server (or WAN link affecting the backup environment.)

 

 

View solution in original post

3 REPLIES 3

RahulG
Level 6
Employee

I guess you have created the backup jobs on the central admin server , it should be created on the managed media server , I am not sure about this as have not used the CASO option for a long time . You go through the following post 

https://www-secure.symantec.com/connect/forums/be2012-wan-optimized-deduplication-communication-or-c...

akuma22
Level 3

But the CAS is at my primary site and so is the data I'd like to backup. Are you saying I shouldn't use the CAS to backup anything locally but instead implement an additional MMS?

Once again, what happens when my CAS becomes unavailable? it will render any MMS useless. I believe I need to replicate some configuration data between both servers but am unsure exactly what and how to do this.

 

Edit: The link above describes my exact scenario. I had hoped backup exec would be somewhat more intelligent but it seems I am left with 2 long winded options after CAS failure or loss of connectivity to CAS:

  • Rebuild and restore CAS
  • Convert MMS to stand alone mode and then perform restore(s)

I think that both options would take time I don't have in the event of a disaster. Surely there's a better way?

Colin_Weaver
Moderator
Moderator
Employee Accredited Certified

The Opt Dup configuration gives two choices.

 

1) Original backup done to MMS/MBE, OptDup to CAS. In which case if CAS (or link to CAS goes down your initial backups may not run as the MBE depends to a certain extent on the CAS database

2) Original Backup done to CAS, OptDeup to MMS/MBE. In this scenario your initial backups should always run, the duplicates will fail if the WAN link or MMS/MBE is unavailable. Again the CAS going unavilable would also kill your backups,

 

In either scenario you could spend time if a failure happens, doing a reconfigure of whichever server is still running back to stand alone mode, which could then be used for Backups or Restores (assuming your backed up data is available. However you would probably be better off spending time fixing the problem and then just re-run the jobs once fixed.

 

Basically this is not a multi-master model so cannot achive full resilience and some manual labour will always be needed to recover from a lost media server (or WAN link affecting the backup environment.)