Hello

I was able to buy following Stuff:

2 x IBM Bladecenter H
6 x IBM HS20
8 x IBM HS22
16 x IBM EXP5000 with a lot of Harddrives
2 x DS5100
1 x IBM Console Switch

including i think everything i will need...Cables, IBM CD's an so on...

My plan is to setup a HPC Cluster with Qemu for Virtualize one big Windows Machine to do some Renderjobs...
I think for that i only will use one Bladecenter and perhaps one EXP5000
The cluster is the right way for me because if i have one powerful computer i dont have to do a lot of updates on every blade. The software i use have a lot of big updates...
At the moment i run a Renderfarm with 15 normal Windows Computers... and that make no Sense because if there is an Update i use a half day to update every Computer...

To realise this i think i would take the solution with Rocks Cluster. It was the onliest solution which i understand at the moment. But because i have no experience in this things i wan't to know what you think about. Are there better solutions for my equipment?

But first i have to Install the Powerplugs for the Bladecenter in my House :-D
I hope its the right Topic in this Forum...

In between my last post i found out its better to share render jobs to a lot of small machines then on one big machines. It means for me that i will setup so many virtual core machines how i have physical cores. If i doo that i would be able to setup the server not as a cluster. Does this bring me any disadvantages?

What do you think about a open stack? Would it help me or is it not necessary for me? I have found this Methode i think i would be able to do it...and it looks to me that i would be much faster with this methode...
https://www.ubuntu.com/download/cloud/autopilot