possible to do two concurrent restores??

[tommls]

We have a hardware refresh coming up...and our two largest servers (email and file servers) will have to be restored.
Is it possible to run two concurrent restores of these two virtual servers?? -- and if so, how do I do this??
(The restore window so far as I know is modal, one can't do anything else while it's open, is it okay to close this window and start another restore job??)
Or would this all be so slow that I'm better off doing one at a time??
Restore speed is something like 60-70 MB/sec and each server is over 600 GB...
Getting these two servers up and running again is the main biggest task of this hardware refresh.
Thank you, Tom

[driley]

Could both the new and existing environments be up at the same time? If so, you could consider using replication to elimnate down time.

[foggy]

Hi Tom, you can close the restore window, it will not interrupt the restore process. As for the parallel restores, it is definitely possible but depends on your infrastructure capabilities, basically backup repository I/O and proxy server(s) specs. Similar to backup, each VM disk restore will be a separate task so the number of concurrent tasks that your proxy (or proxies) can handle plays the role here.

[fazzauk]

In v10 you can, not so sure about earlier versions

[tommls]

Hi Tom, you can close the restore window, it will not interrupt the restore process. As for the parallel restores, it is definitely possible but depends on your infrastructure capabilities, basically backup repository I/O and proxy server(s) specs. Similar to backup, each VM disk restore will be a separate task so the number of concurrent tasks that your proxy (or proxies) can handle plays the role here.

How and where do I look up about the concurrent tasks our proxy can perform??
We have an all-in-one Veeam server to which I can temporarily add more vCPU and RAM.
I know we can have at least two backups running concurrently, since our backup jobs do overlap, with multiple VMs within each backup job.

Thank you, Tom

[foggy]

Concurrent tasks are configured in the proxy/repository settings. Keep in mind that for all-in-one installation both components are running on the same server, so the load adds up correspondingly.

[tommls]

OIC -- our proxy is set for 8 concurrent tasks, we have 12 vCPU and 24 GB RAM allocated, I will make it 32 GB RAM temporarily and that should make things work okay.
Wrt restores does the quantity or size of disk volumes make any difference??
I ask because the file server has one small C: drive and one very large D: drive, while the mail server has several volumes of various sizes.
Thank you, Tom

[foggy]

Each disk will take a single task slot regardless of its size.

[tommls]

The two servers involved have 9 disks total.
Will Veeam refuse to do one of the restores, or will Veeam start processing the 9th disk as soon as one of the first 8 disks are finished??
If the latter, I should start the mail server first because some of its disks are small, wait 10-15 minutes, then start the next restore of the file server.
Thank you, Tom

[foggy]

The disks that do not get slots will be queued up and wait for resources to be released after processing of other disks. In fact I'd start the file server first to allow the large disk to get a slot from the very beginning and then the mail server will get the slots for all its disks once the system disk of the file server is done. And no need to wait 10-15 minutes, you can start them with a shorter interval, tasks will be queued appropriately. Thus you'll optimize the restore window.

[tommls]

Oh, that is so sensible. Thank you, I will just only temporarily add some RAM to ensure there's no unforeseen hiccups.