SureBackup - do you use this?

[ddockter]

I'm curious how many are using SureBackup on a regular basis to check their backups? I tried using it with one of the earlier versions of Veeam, but it just seemed that most of the SureBackups would fail for one reason or another so it just didn't seem to be worth the hassle. If you are using it, do you run SureBackup jobs after every night's backups or just periodically? Do the SureBackups have much of a performance impact on your vSphere environment?

[vmexpert]

We started using this a few months ago, after our first recovery problem with full VM restore of our critical application server. The VM could not start, so we restored it, but the same issue was still present. Long story short, it turned out VMX file was messed up before the backup was taken. Took us many hours of troubleshooting and a support case with Veeam to figure out the issue was on our side. This has completely blown our SLA, as recovery time went through the roof. Obviously, SureBackup would have caught the issue, so we are now validating all important VMs every night, because I feel the heads will roll if something like this happens again

We start SureBackup right after our backup jobs complete, so there is no impact on our vSphere environment as nothing happens at this time (late night). But I suppose if I happened to be running some severely overprovisioned 24/7 shop, then having went through the hell above once, I would just assign a dedicated host for this activity specifically, no big deal (just some old unused server).

[foggy]

Doug, depending on the environment, SureBackup could indeed require thorough initial setup, but once configured, can save tons of work ensuring your backups are always recoverable. Anyway, you can always ask for assistance here on forums or contact support directly if you experience issues with creating virtual labs/running SureBackup jobs.

[Gostev]

most of the SureBackups would fail for one reason or another so it just didn't seem to be worth the hassle

Unless your backup storage is really slow, which is a common reason for failures, this is actually quite alarming, as this means you will get the same failures "for one reason or another" when recovering from real disaster.

[ddockter]

Almost all of the VMs fail to boot within the allotted time of 600 seconds. My repository is a Exagrid dedupe appliance.

[Vitaliy S.]

Hi Doug, I believe I can explain this behavior with your SureBackup jobs. See this existing topic for further reading > Surebackup with Exagrid.

Switching to a regular storage device (without dedupe) will most likely resolve all your timeout issues.

[ddockter]

Unfortunately I'll be using the Exagrid as my repository for awhile as I do not have available space elsewhere for my backup repository nor do I have funds available to replace the Exagrid.

[Gostev]

Couple of comments:

1. If you size the landing zone in ExaGrid accordingly, you will be able to run SureBackup of the latest restore point at full speed of raw storage, and avoid the issues you are facing today.

2. I hope what you said does not mean you maintain a single copy of your backups on ExaGrid alone, as one copy is never enough.

Normally, I don't recommend backing up directly to deduplicating storage (performance considerations), but rather use it for what it was designed (efficient data archiving). I recommend that you backup to the fast primary backup storage, and maintain just a handful of restore points there (for operation restores, SureBackup etc.). You can then copy those backups with Backup Copy jobs to secondary, slower storage with lower cost per TB (e.g. deduplicating storage) and maintaining more restore points there to meet the required retention policy. This is also they cheapest way to meet the requirement of maintaining at least 2 independent copies of your backups.

For example, you can build your primary backup storage from an old Windows or Linux server with just a bunch of internal disks, this makes it perfect primary backup target really cheap. You should certainly be able to afford this, if you could afford buying ExaGrid?

[ddockter]

I will have to give this some consideration. I have 4 open drive bays in my Veeam Proxy server (Windows 2008). I could potentially add drives and put my repository there and then do Backup Copy jobs to the Exagrid as you suggested; or would this adversely affect my backup performance? Also FYI I currently copy the backups off to tape via Backup Exec and will continue to do so even if I implement the above.

[foggy]

I could potentially add drives and put my repository there and then do Backup Copy jobs to the Exagrid as you suggested; or would this adversely affect my backup performance?

It will not, as backup copying normally happens outside of the backup window.