The introductory training documents that Intel provides with the Cluster Studio XE Suite want/expect the nodes of the cluster to be configured with the Intel provided mpd ring. We use the Adaptive Computing resource manager Moab with the Torque hardware manager.
Do you have of any good source of documentation for getting the Intel Cluster Suite to play nicely with the Moab/Torque queue manager?
As discussed via email, we currently do not have any significant documentation for integrating with Moab*/Torque*, as most of the integration happens automatically. When mpirun is invoked, it will check the environment to determine if it is running inside of a Torque* environment. If so, it will inherit the node list from Torque*.
We are examining options for documenting our job scheduler integration.
Technical Consulting Engineer
Intel® Cluster Tools