Hi
We at our company thought of implementing "Oscar"(open source for HPC needs)
I did make some reading and will be happy if somebody in this forum have experience with or understand the "Oscar" application better then me and can reply.
My boss and I thought that with "Oscar" we need to install one machine with Linux lets say in one server and then connect thought local communication switch or hub the other nodes(as much more as better)those over all cluster will act as one big large pool of Cpu and memory at the top of this big cluster we can install Linux that will be the up level OS that when I will do : "cat /proc/cpuinfo" I will see 20 Processors or maybe "cat /proc/memiinfo" 40GB memory (from this above layer i can use all the pull of the Cpu or memory or storage as for my needs (like vmware)
Is the above theory is correct?
Can I do with "Oscar" this action?
Can I use Oscar as the underlay to use my hardware resources as one big pool?
Thank in advance for any help!
We at our company thought of implementing "Oscar"(open source for HPC needs)
I did make some reading and will be happy if somebody in this forum have experience with or understand the "Oscar" application better then me and can reply.
My boss and I thought that with "Oscar" we need to install one machine with Linux lets say in one server and then connect thought local communication switch or hub the other nodes(as much more as better)those over all cluster will act as one big large pool of Cpu and memory at the top of this big cluster we can install Linux that will be the up level OS that when I will do : "cat /proc/cpuinfo" I will see 20 Processors or maybe "cat /proc/memiinfo" 40GB memory (from this above layer i can use all the pull of the Cpu or memory or storage as for my needs (like vmware)
Is the above theory is correct?
Can I do with "Oscar" this action?
Can I use Oscar as the underlay to use my hardware resources as one big pool?
Thank in advance for any help!