Video Screencast Help
Symantec to Separate Into Two Focused, Industry-Leading Technology Companies. Learn more.
Archiving and eDiscovery Community Blog

“Thinking About Disaster Recovery?”

Created: 13 Mar 2009 • Updated: 29 May 2014 • 5 comments
Dano Oliveira's picture
+3 3 Votes
Login to vote
As a product manager for Backup Exec I have the pleasure of talking to many of our customers and partners in the field. It’s fascinating to hear about the latest backup strategy, with the newest hardware and the way customers use our product to secure good backups. On the other hand, what concerns me is the lack of strategy when it comes to disaster recovery.
 
When I did IT work, which wasn’t too long ago, I remember having to rebuild a server from scratch when it crashed. True, I had my data secure, but the time it took to rebuild that server cost the customer or company money. It was very painful; I remember having to locate the drivers disc, install the OS again, get everything updated, reinstall the apps, configure everything, etc… and then restore the data.   After that, I had to make sure everything was synced and then reconfigure the backup job.
 
Of course this is just the Reader’s Digest version. You, who have gone through this, know I didn’t cover this in its entirety. Needless to say, it’s quite time consuming and it could impact productivity, revenue, and moral. In some cases, it could have an impact on your position in the company if the proper disaster recovery plan is not in place.
 
Questions for you to consider:
  • Are you prepared?
  •  If your Exchange server went down today how long would it take to get your system back online? 
  • What is your RTO & RPO and do you meet those objectives?
 
I’d like to hear about your Disaster Recovery strategies. Share some glory stories and feel free to share some horror stories. If Backup Exec System Recovery saved you, please share.

Comments 5 CommentsJump to latest comment

Swathi Turlapaty's picture

Dano, it's a great call from u asking users to share their disaster recovery success and horror stories. Folks, here is another option for you to share your experiences with your peers.

Thanks,

Swathi

+1
Login to vote
ITSpoint's picture

Need to have an idea of what utility to use with Veritas to plan disaster recovery of my network using backup exec 11d that i already have.

In the past i have recoverd missing or deleted data for users but fortunately not had issues with mail store or recovery.

I have just upgraded my servers and migrated from win 2003 Exchange to 2007 running on win 2008 and on a win 2003 domain.

My recovery plan is to have a server off site to enable recovery of the system in the event of disaster. I presently do not have a good RTO as we only take the tape backup of the network off site and have no clear recovery plan.

What do i need as a best practice to get a recovery plan with a resonable RTO.

+1
Login to vote
Dano Oliveira's picture

 

ITSpoint,
It’s good that you bring up best practices. We have two products for a combined solution that we would recommend to create the best RTO & RPO for complete disaster recovery.
 
Backup Exec 12.5 w/ Backup Exec System Recovery 8.5
·         Backup Exec 12.5 for your RPO
·         Backup Exec System Recovery 8.5 for your RTO
Benefit of a combined solution
Set the right Data Protection strategy with Backup Exec 12.5?
Data changes frequently so schedule your backups according to your need (Continuous, multiple times in a day, daily)
 
•       More frequent file/application level backups are needed
•       Allows for more granular file recovery
•       Most commonly used for document and email recovery
 
Set the right System Protection strategy with Backup Exec System Recovery 8.5?
Systems changes are intermittent so create a recovery point when those changes occur like; OS Service Pack, installing a new application, applying a patch
 
•       Allows for point-in-time, full server recovery and patch rollback
•       For some servers, customers may prefer the system level, everything-at-once approach to backup
•       Most commonly used for disaster recovery and migration
 
Here are a couple of Use Scenarios for you:
Solution Scenario 1
•       BEWS is used to protect the server resources in the environment
•       BESR is used to protect the BEWS media server
Solution Scenario 2
•       BEWS is used to protect data and applications on server resources in a customer’s environment
•       BESR installed to critical servers for fast system recovery
Solution Scenario 3
•       BEWS is used to protect data and applications on server resources in a customer’s environment
•       BESR is installed to critical servers for fast system recovery; also converts backups to virtual format and stores on hypervisor
Solution Scenario 4
•       BESR is used as primary backup solution for important server resources
•       BEWS is used to store BESR recovery points to tape (offsite and long-term archival)
 
 
Let me know if this helps.
+1
Login to vote
Ryan Shaffer's picture

Hi, Dano.

Its too bad you haven't gotten many responses.  I'd be interested to hear what other organizations are doing.  The solution we are implementing for my organization is probably a hybrid between solutions two and three you shared above.

We currently use BEWS 12.5 for backups, which are written to tape and rotated offsite weekly.  We want to reduce RPO since the potential data loss could be up to 1 week.  Additionally, since our infrastructure is fairly complex for a smaller organization, recovering from a disaster would require considerable time.  We are hoping that BESR will help us eliminate the time and complexity of recovery.  Our early testing has us very excited about the possibilities.

I'd be curious to hear what advice you or others have regarding the following solution:

  • BESRM manages all BESR instances.  
  • All servers except the BESRM server are scheduled in a recovery point set (weekly).  
  • Recovery points are stored on a volume of the BESRM server.
  • BESRM server is BESR'd once all other recovery points are complete.  The end result is a single recovery point that contains our entire environment (that is selected for BESR).
  • The "Master" snapshot would then be backed up to tape nighly.  Is there any advantage to this over backing up each recovery point set individually?  We thought bringing up the entire backup server would make it easier to facilitate recovering other servers.
  • We would still use BEWS for data restoration.
  • I should also note, we do a conversion job on our domain controller to create a virtual machine.  The idea is to bring this up first, then the backup server and recovery points.  Then our other servers based upon their priority.

Is this too much?  Anything I am missing, not thinking about?  Early estimates place the size of the recovery points for all BESR'd servers at approcimately 500GB.  We use LTO3 tapes and can write up to 800gb of compressed data per tape. 

We still need a workable solution to reduce the RPO in the event of a site disaster.  Online backup is attractive, but expensive.  Maybe it is as simple as rotating an external hard drive off site daily, or daily exchange of our tapes.

Thanks,

Ryan

0
Login to vote
Sheena K.'s picture

@ Ryan Shaffer

We are doing something similar the the BESRM. We rotate it offsite every week. We are contimpating and hoping to move to doing the "master" once a week and then doing a "modified" update to be appended to the SRM daily so that at any point we would have a full data backup and be able to restore quickly.  We have discovered so far that there does seem to be a limit of using the SRM of extremely large data servers. One of our data servers is over 4tb and we are unable to complete a "master" SRM in a weekend.

As another part of ourt backups and disaster recovery we do daily online, tape and offsite backups. The offsite backups are online backups to one of our other offices. The only thing about online and tape backups that we discovered durning a Disaster Recovery test was the file character limit that seems to be a new limit imposed on BE12.5.  We were initially doing our  data server "master" tapes online then compressing and encypting them to tape for offsite storage. However we discovered when trying to restore the online backups they would fail, claiming not all the media had been inventoried. After calling Symantec we discovered that the character limit was causing the online backups to become corrupt. We now have to do the "master" backups straight to tape until we can find another solution.

We are in the process of examining our whole backup and recovery process so I would love to hear how other company's are doing their backup and their process for recovering.

Best,
Sheena K.

0
Login to vote