Adding more VMs to existing backup (and replication) jobs

[dsimpkins]

I have about 30 VMs, and because the people who implemented Veeam for me didn't adequately explain themselves or the Veeam product to me, I created 30 backup jobs (and replication jobs).

Since troubleshooting and exploring Veeam more and more, im trying to conform more to Veeam's recommendation of combining similar VMs into the same jobs. For example, i have a server called AV01 being backed up within a job called AV01, and then i have another VM called RAS01 backing up within a job called RAS01. What I have done is disabled the RAS01 job, and added RAS01 VM to the AV01 job, because coincidentally they were both running at the same time, to the same datastore, with exactly the same job features such as restore points etc.

I've done this with many of my VMs now, however my simple question is: do i need to hold onto the old backup restore points? Let's take a daily backup job with a 7 day retention period. The reverse increment VRB files are reflecting the changed job name (called "AV01, RAS01") however the largest file, the VBK file, still has it's original name (AV01).

- Can someone provide some advice as to the best way to tidy this up?
- Im wondering whether I can simply delete the files associated with the redundant jobs, and then perhaps i should then run a "full" backup on the newly modified job with the additional VMs.
- Does this affect replication jobs? Also, some advice as to how i start combining replica jobs would be good too - im replicating over a WAN so i assume the dedupe process happens at the source first and combining jobs doesn't really effect the target replica VMs but instead reduces the amount of traffic being sent.

Thanks heaps

Daniel

[dellock6]

Hi Daniel,
for sure grouping several VMs in the same job is a better way to deal with backups: deduplication for example is made at the job level, so the more VMs Veeam can compare, the better dedup ratio you can gain. Instead of dealing with 30 jobs and their schedule, you can use proxy and repository limits about cuncurrent jobs execution (based on the performances of those components) and then let the job be queued automatically by Veeam. Even starting all of them at the same time.

About you other questions:
- backup file name reflects those of the job, I think you need to run an active full to change the file name after you changed the job name
- you can delete old backup files after you have enough restore points in the new ones
- replication jobs and backup jobs are different and have no interaction apart working on the same VMs, you can schedule them as well since Veeam will operate on a given VM only once, so if a replica job needs to work on a VM that's is beeing open by a backup job, replica job will be queued.
- to reduce replica traffic, be sure you have at least two proxies (one at source and one at destination) and you have configured replica job accordingly

Luca.

[foggy]

Daniel, after the first run of the combined job VBK file will contain both VMs, while VRB files will still have only one of them. So you can keep the old restore points from the second (disabled) job until you have the required number of restore points for the new one. Old backup points of the active job will go away according to the retention settings. For the name of the VBK file to be renamed, active full is required as this will re-create the whole VBK.

Replica restore points are stored in a native format (as VMware snapshots) and are not compressed and deduplicated. Some sort of dedupe is performed on the source-side only (the content of virtual disks is consolidated, overlapping blocks, zero blocks and swap file blocks are filtered out). So combining VMs in one job does not affect replication.

[dsimpkins]

thanks guys (interested in this concept of starting all jobs at the same time!). I have enough restore points in VBR files (of which the name reflects the new VM), but i want to know if the VBM file would contain the full backup should i need to restore the whole VM. Or do i have to do a full backup first.
If it helps, here's the files in question:

AV01+RAS012012-10-02T010028.vrb (402MB)
AV01+RAS012012-10-03T010028.vrb (481MB)
AV01+RAS012012-10-04T010035.vrb (384MB)
AV01+RAS012012-10-05T010037.vrb (422MB)
AV01+RAS012012-10-06T010039.vrb (365MB)
AV01+RAS012012-10-07T010022.vrb (365MB)
AV01.vbm
AV012012-10-08T010028.vrb (10.2GB)

Does the file "AV012012-10-08T010028.vrb" now include BOTH full VMs, or is it only the other 6 files that contains ONLY those day's increments, and if i wanted to restore a full RAS01 VM, then i would not be able to?

Thanks

[foggy]

AV012012-10-08T010028.vrb (10.2GB)

This should be AV012012-10-08T010028.vbk, not .vrb. This is the full backup containing both VMs starting from the first run of the combined job, as I said above.

[dsimpkins]

correct, vbk, sorry i copied & pasted.

and sorry, Alex, I just wanted to clarify my question and make sure the answer was the same. Thanks for the reply - sorry for the bother.