Other Tech

I thought I would post on this feature because it seems it is often overlooked and not everyone is aware it exists. This feature is the Global Exclusions Selection List in Backup Exec and it has been around for a very long time.

This global selection list allows you to exclude a path or file just once and this exclusion will automatically apply to every job you create. This is a lot easier than deselecting the same checkbox for a common exclusion across dozens of servers and jobs.

This came to light recently because I had to add global exclusions for the Symantec Endpoint Protection 12.1 client, which kept causing dozens of exceptions in my job and I really didn’t need to recover this program data in the event of a catastrophic event. I could simply reinstall the client if needed.

I know most of my posts have been on Backup Exec 2012, and I know a lot of people are still using Backup Exec 2010, so I thought I would design this article for setting this up in Backup Exec 2010.

PART I: http://supertekboy.com/2013/01/12/symantec-backup-exec-to-include-job-based-backups-maybe/

Back in January, I wrote a post about the possibility of job-centric options coming back to the Backup Exec product line. My inside sources had told me October initially, then later February and now I guess later still. For more details on the job-centric versus server-centric debate please see the previous post.

One thing I wanted to add to the argument against, server-centric backups in 2012, which dawned on me recently, was the use of an autoloader or robotic tape library with Backup Exec 2012.

Going back to the previous version of Backup Exec, 2010 and earlier, and using my 10 servers example from the last post, most people are aware that the average robotic arm takes 2-3 minutes to move a tape from a slot and into the drive. The same can be said on the back end of the job where the tape is removed from the drive and moved back into the slot.

Under the job-centric philosophy of 2010, one job that backs up 10 servers, only had this movement happen twice. Once at the beginning of the job and once and the end of the job (unless your job spanned multiple tapes – for the sake of this example lets assume you can fit it all onto one tape).

Under the Backup Exec 2012 server-centric philosophy, those ten servers are now ten individual jobs. That tape now moves twice for each server/job. So, the same backup now has 20 tape movements in a robotic library/autoloader as opposed to 2.

The problem here is the amount of time that adds to the backup window you used to have in 2010. In 2010 the average robotic-libraries tape operations were likely 4-6 minutes. Now using our 10 server example in 2012, the tape operations have been multiplied ten-fold, so your tape operations have increased to 36-54 minutes. Adding 54 minutes to your backup window may not be desirable.

Hopefully, we will see job-centric return soon as an option.

For any of you that own a Backup Exec 3600 appliance, you already know that it is a locked-down version of Windows Storage Server 2008. Console access (through a keyboard, mouse, or monitor) is not recommended after the initial setup, and that Remote Desktop is turned off. The only method of administration is through the appliance web portal.

For those of you that are looking for a little more, you can enable RDP through PowerShell. Keep in mind that enabling Remote Desktop is not supported by Symantec and is not considered a best practice. Do so at your own risk.

I have seen this issue a few times now while performing a disaster recovery of a Backup Exec 3600 appliance. There are many DR scenarios for a Backup Exec 3600 appliance but the one I am referencing in this post is when the appliance is a total loss and you need to move your four data drives from the old appliance to a new appliance.

I have seen some instances that after you move the 4 existing data drives from the old appliance to the new appliance, you get past the Symantec POST screen only to be left at a white cursor that is blinking in the top-left of a black screen.

I ran into an issue recently where I could not do a GRT (file & folder) based restore from a backup that had been created using the VMWare Agent. The backups all reported as successful and I confirmed the GRT settings were configured in the backup options correctly. I also confirmed that the Remote Agent was deployed on the virtual machine and communicating correctly.

Basically, when I selected the virtual machine in Backup Exec 2012 (because you do a GRT based restore through the Remote Agent) the restore wizard only had the option for a full virtual machine restore. There were no other options in the restore wizard. I could only restore the entire virtual machine and all its virtual disks.

Apparently this is a known issue but there is a very simple workaround. To be able to do a GRT based restore from a backup using the VMWare agent I had to first do a one-time backup through the remote agent, treating the virtual machine as if it were a physical box. It didn’t need to be much. In fact, I just backed up one, yes just one, tiny little file through the remote agent. It didn’t even matter what destination I used. Once the backup completed I went back into the restore wizard and Hey! Presto; I now had an option to restore files and folders. Furthermore, it showed every single file and folder from the full virtual machine VMDK backup.

(At the time of writing there is currently no hotfix.)