Using Virtual Proxy to take advantage of spare resources?

[v334m]

So i have 3 physical servers i'm using as proxy and repository roles, however i had a thought i wanted to bounce off the community...

We dont pay for CPU on demand like you would with AWS etc, we have dedicated 28 cores and many TBs of RAM...
Over night (when all the backups run) the overall CPU usage of our cluster goes from about 50% (day time) down to <25% between say 19.00 and 07.00

Could i use a VM to utilise these spare resources? We have several hundred VMs to backup (and a tonne of data) and extra proxys would really help for running many concurrent tasks (lots of vms, with lots of disks)
Would love to hear your thoughts on this, and any tips on how many CPUs to add.

Obviously the jobs using this proxy would have backup windows (as all my jobs do) not to run during business hours, meaning the proxy VM would be idle throughout the day, releasing the resources back into production.

Cheers

[foggy]

Yes, having additional virtual proxy servers will allow you to run more tasks concurrently (and use hotadd for restores, should something goes wrong), just make sure your repositories are capable of accepting the load. You can find system requirements for the proxies in the corresponding user guide section, but the rule of thumb is a single core and additional 200MB of RAM per each task.

[v334m]

Thanks for the quick reply, single core - setting up how many concurrent tasks - just two? Or because of being virtual, we can add more depending on back-end resources?

Many thanks

[foggy]

Each task requires a single core, you can add as many cores as you need to run the required number of concurrent tasks.

[v334m]

Great, so in theory could add 8 or more cores, providing the vm can get the resources quick enough, would be nice to test it out.