I was wondering whether the following stack would work, and if so then how well, and what sort of problems I might expect to encounter when setting it up?
- Hardware layer - lots of cheap servers
- SMP layer - ScaleMP
- OS - Linux 64-bit (e.g. Red Hat Enterprise Linux 6, or any supported by ScaleMP)
- VM host - VirtualBox or similar running on Linux
- VM OS - Windows Server (insert relevant edition here)
- Resource-hungry applications running on Windows Server
What I hope would happen here is that I'd be able to create a number of windows VMs that would each be able to utilise CPUs and memory from across the whole pool of server hardware. This is of interest to me, as the applications in question are Windows-only.
As far as I can tell, each layer is compatible with the next, but I've never tried to set this up before and I don't have suitable hardware to test it on right now. I expect a fair bit of overhead to run a cluster like this, but I'm not sure how much. Any advice appreciated.