- Posts: 41
- Liked: never
- Joined: Oct 06, 2010 6:54 pm
I am new to Veeam. We are running Veeam 5.0 and I just have a concern regarding Virtual Lab.
I’d like to implement Virtual Lab on one of our Production host, because I like to have a capability to restore VM instantly from the back up and run in virtual lab temporary in case we have couple of our Production VM’s crashed.
My concerns is am I going to have an issue to run virtual lab on one of our production host on the same network?
Any best practice suggestions would be highly appreciated.
For the case you described, Instant VM Recovery is the closest convenient feature to use. VMs are powered on directly from backup, thanks to vPower technology. Then, you can use SVMotion to move 'em to production datastore. Please keep in mind that running directly from backup is not so fast in terms of IOPS (because backup file is compressed and deduped). This method can be treated as "spare tire" and is not recommended for permanent usage.
Also, you should consider that keeping these VMs powered on is ok only if they are not plugged to network (there is a corresponding checkbox). Otherwise there might be a conflict with production VMs running on the same subnet.
You are talking about 'virtual lab', but using surebackup is not very applicable for your purposes: VMs running in lab get slightly tuned by B&R (for ex., network mapping is changed - for isolation purposes). So, using virtual labs is not for your goals.
In conclusion, I would recommend to play with Instant VM Recovery and decide whether performance and resource consumption is ok for you.
Hope this helps.
- SVP, Product Management
- Posts: 29918
- Liked: 5852 times
- Joined: Jan 01, 2006 1:01 am
- Location: Baar, Switzerland
You also seem to mix up 2 different features together. Running VM in production in place of crashed VM is what we call "Instant VM Recovery" (this does not use virtual lab at all, as explained in FAQ).
On the other hand, virtual lab is used for all other use cases (SureBackup, U-AIR, Sandbox) - this provides for completely isolated environment to run VMs in, so that you can do this in parallel with running the same VM in production, and without affecting the latter. This is exactly why virtual lab cannot really be used with the use case you have described - running some VMs in place of crashed ones.
I highly recomment that you review the sticky FAQ topic, as it goes in details on both of these use cases, and so should remove the confusion you currently have. It also provides best practice suggestions you are looking for.
Also, if you are completely new to the product, I recommend signing up for live demo with one of our engineers - they will be able to provide you with good technology introduction and overview.