none
Issues with template properties for Max Cores Per Node RRS feed

  • Question

  • Hey guys, so we have a very large model we are trying to get to run.  The issue is it needs lots of HDD space so i calculated if i isolate this model to 9 servers(trying to use 71 cores) and use the template properties  for max cores per node set to 8 it would use the list of 9 servers. While limiting it to 8 cores on each, so they would not run out of disk space.  This has not worked. Submitting to the job template setup this way, errors out with: "This job requires as least 1 Cores, but the list of candidate nodes that the job scheduler service returned for this job contains only 0 cores"  

    It seems like if the max cores per node is lower than the cores request by the job the scheduler cancels the job. Why?  I have 9 servers selected with requested nodes and limited to 8 cores per node should only be 72 cores, which would be available spread across 9 servers.

    Monday, July 15, 2019 4:56 PM

All replies

  • Hi kenjar0497,

    The "Maximum Cores Per Node" in a job template means to choose the nodes whose number of cores is not larger than that value to run the job. It doesn't mean to use only the number of cores on the nodes. You may try to undersubscribe the number of cores on the nodes if you want to limit the number of tasks to a max number that is less than the actual number of cores on node.

    Regards,

    Yutong Sun

    Friday, July 26, 2019 4:16 PM
    Moderator