New Veeam Deployment

[borgiaj]

Hello everyone.

I'm new with managing Veeam and come from mainly a traditional disk-to-tape or disk-to-disk-to tape background. I've taken a bunch of the online training, watched various youtube videos, and I'm trying to wrap my head around things. We have a very small VMware environment - 14 Linux/UNIX systems and 11 Windows systems. So far, I've managed to get this up and running, and things seem to be going well, likely because the environment is so small. We're not doing anything overly complicated with Veeam at this point other than backing up the VMs and having them available for file-level recoveries and such if needed. No replication or anything (yet). The Veeam server itself is a Server 2016 VM. All of the VMs are backed up to large local disks attached to the Veeam server. I have one large file server that we're doing tape jobs on, because that is the customer requirement as well, also for the sake of space on the array we're backing up to. I guess I have some very basic questions.

Backup Proxy - my understanding of this is that it offloads work from the main Veeam server for processing. Do they require their own backup repositories to shoulder the load of storage. The directly attached volumes to the Veeam server VM makes that particular VM huge. Would I even need proxies for backing up such a small number of VMs?

Service Account - most of everything I have setup with Veeam has been setup with my personal administrator account in AD, which just seems to be not the best way. Are people using service accounts to manage their Veeam environments?

Veeam server VM or physical server - sometimes I wonder if it wouldn't make more sense to run the actual Veeam server as a physical server, to keep it separate from the virtual environment. I guess my reasoning for this is, it would be problematic to try to evac that VM from the storage array and blade server where it resides in this building to the storage array and blade server in our other building, because it's so huge due to the local storage repositories. Is there a better way to manage these storage repositories?

Overall, I've got to say, I really like this product. Recovering VMs is very fast. File recoveries seem to work perfectly. Backups are FAST. Like I said, I've got it running, and it seems to be fine, but I just feel like I could be doing this better.

I'm sure I'll have more questions, but it'd be great if someone could point me in the right direction with regard to the above questions. Any suggestions would be appreciated.

Thanks,
Joe

[Rick.Vanover]

Hi Joe, welcome.

A few perspectives here from me:

Backup Proxy: Yes, it does the data movement work on behalf of the Veeam Backup & Replication server/console. You can separate proxies from the server to distribute the work, it's a Windows service. You don't need a repository for each proxy; though if you have a proxy and repository on the same system (such as a physical server with a lot of drive space), you can have nice efficient transfers.

Service Accounts: I don't like using the same account for all tasks, with the communication to backup repository being the most critical. In today's ransomware era, I recommend putting in as many explicit accounts and permissions as possible. These are all well-documented here: https://helpcenter.veeam.com/docs/backu ... l?ver=95u4

VM vs. Physical: Oh this question... It comes up a lot, it's been discussed here often, but I think if you put the proxy physical (as well as the repository outside of the VM environment) that is the more important topic. I like having the proxies physical as during the backup window, it has a spike in CPU/RAM usage; it's nice to have that outside of your VM environment. So yes, you are on to a good preference as you indicate for physical for the Veeam server. The best way to 'manage' the storage repositories is to use the Scale-Out Backup Repository, it is a logical construct composed of many backup repositories. Nice for evacuations, migrations, etc.

Does any of this help?

[borgiaj]

So it DOES make sense to split out the backup repositories using proxies to keep the Veeam B&R console server separated out (and a manageable size). The B&R console sitting on a VM would be the brains of the operation, while physical proxies connected to large pools of disk for repositories would do all the work. That's what I'm seeing in my head. Even though we're small in size, maybe just two proxies would make sense to split the load. What do you think?

As far as the service accounts and permissions, I guess I'm going to have to go back and take a look at how I'm handling that after re-reviewing that link you sent. Thanks.

I was recently looking into scale-out backup repositories. In this case, would both proxy servers point at the same scale-out backup repository? Seems like that's what I'd have.

I really appreciate your advice on this stuff.

[foggy]

If you're feeling like you're close to the limits of your backup window, then spreading the load makes sense, indeed. Yes, both proxies can work with the same scale-out repository as a target.