At my university we are adding centralized computing resources (blade servers) that need to be split/shared between 2 purposes:
1) renderfarm managed by Qube (already purchased)
2) hpc compute-cluster managed by LSF (or something similar?)
Nothing is in production as of yet and there is more flexibility on the cluster computing side. The nodes will be Linux-based and GPU/CUDA capable.
As new services, their relative popularity and usage pattern are yet unknown -- I'm wondering what options are available to allocate the nodes dynamically. One strategy is, of course, to assign a separate/fixed set of compute nodes to each manager and adjust the balance over time. Another approach might use Qube's ability to (un)block worker nodes on a schedule with a corresponding action on the LSF side (?).
I have inquired directly to pipelineFX about this as well. I have heard (indirectly) that some level of integration is possible, but nothing concrete so far...
Your input is much appreciated -- many thanks!