B2D > D2D > D2T?

Hello Lee,  

With the kind of requirement, you have, I'd suggest that you make use of the Optimized Deduplication feature with Backup Exec 2012. 

You may refer to the following links to get more details on it: 

Best practices for using Backup Exec 2012 Deduplication Option with the Central Admin Server Option: http://www.symantec.com/business/support/index?page=content&id=HOWTO74447

How to do manual seeding before running a Optimized Duplication Job: http://www.symantec.com/business/support/index?page=content&id=TECH163110

 

Hi,

If you're running anything other than BE 2012, you need CASO. If you have BE 2012, then you'd license the Enterprise Server Option (ESO which includes this).

There would be no problems running the CASO server on a separate server with disk and tape attached to it.

If you're looking at using dedupe you'd need to use optimized dedupe.

Check the links below:

http://www.symantec.com/docs/TECH43297

http://www.symantec.com/docs/HOWTO73626

Thanks!

Thankyou both for the advice. Will look into the dedupe option further.

A few more questions:

"There would be no problems running the CASO server on a separate server with disk and tape attached to it."

- Just to confirm, the disk-based storage would be in Site A and connected to that MMS, and the tape library would be in Site B and connected to that MMS. The CASO would be in a separate site altogether (Site C i guess!). Would that work as I expect it? I assume the entire job would have to sit up at the CASO server in order to see both the disk based storage and the tape library?

- What's the correct procedure for adding an existing standalone Backup Exec Server to the CASO server? The BE admin guide says to do this from the standalone server:

 "Click the Backup Exec button, select Configuration and Settings, and then select Set Central Administration Server."

The problem with that, is that the "Set Central Administration Server" option doesn't exist...

 Thanks,

Lee

Hi Lee,

Ideally you want to be in 1 of 2 scenarios:

1. Tape library attached directly to each server to duplicate backups too after backing up to disk. 1 server can still be a CASO but can fulfill a role as a management/reporting server (or no CASO at all);

2. Have disk-based storage attached to the server with the tape library so that you can backup the remote site's server to a dedupe folder, and then using the CASO and optimized dedupe, duplicate that data to the CASO with disk attached to it, and then backup to tape.

Backing up Site A to disk and then attempting to backup that dedupe folder to tape isn't going to work unless you have a seriously fast connection. Data is rehydrated when duplicating a dedupe folder's data to tape and in effect it ends up the same size as originally backed up.
 

Thanks!

From my estimations it would take 11.5 days to get a full backup from Site A to Site B - without the aid of dedupe, so we would definitely need to look at incorporating the optmized dedupe with these backups.

Example 1 which you just described is what we have currently, with no CASO. But we would like to get rid of tapes altogether at Site A if we can, and only keep them at Site B.

When you say data is rehydrated when writing to tape, I understand that, but where is that done? Does it mean that it will go back to the source to pull the original, non-deduplicated data? (effectively pulling the full amount over the WAN?)

I'm a bit confused about example 2, i think i'm not understanding it correctly. Are you suggesting that we would have no equipment in Site A. Site B would have disk based storage, and the backups would go directly to there from Site A. Then the data is duplicated to Site C (which is the CASO server) to another disk based storage and then duplicated to tape. If so, wouldn't that suffer from the same issue of deduplicated data being written to tape and needing to be rehydrated?

Thanks again,

Lee