Restricting job size requested by parallel jobs in a queue

When users submit, modify, or switch parallel jobs with the bsub and bmod -n option to explicitly request a job slot size, or with the -R option to specify resource requirements, administrators can restrict the number of job slots that are requested for the queue.

About this task

LSF rejects job submission or pends existing jobs that request job slot sizes that are not in this list. LSF also rejects jobs that request multiple job slot sizes. The first slot size in this list is the default job size, which is the job size that is assigned to jobs that do not explicitly request a job size. The rest of the list can be defined in any order.

For example, if the job size list for the queue1 queue allows 2, 4, 8, and 16 job slots, and you submit a parallel job that requests 10 job slots in this queue (bsub -q queue1 -n 10 ...), that job is rejected because the job size of 10 is not explicitly allowed in the list. To assign a default job size of 4, specify 4 as the first value in the list. Job submissions that do not use -n are automatically assigned a job size of 4.

When you use resource requirements to specify job slot size, the request must specify a single fixed number of job slots and not multiple values or a range of values:

  • When you use compound resource requirements with the -n and -R options, make sure that the compound resource requirement matches the -n value, which must match a value in the job size list.
  • When you use compound resource requirements without -n, the compound resource requirement must imply a fixed number of job slots. The implied total number of job slots must match a value in the job size list.
  • When you use alternative resource requirements, each of the alternatives must request a fixed number of slots, and all alternative values must match the values in the job size list.

Procedure

  1. Log on as root or the LSF administrator on any host in the cluster.
  2. Edit lsb.queues, and define the JOB_SIZE_LIST parameter in specific queues.

    JOB_SIZE_LIST=default_size [size ...]

    Begin Queue 
QUEUE_NAME = queue1 
... 
JOB_SIZE_LIST=4 2 8 16 
... 
End Queue
  3. Save the changes to lsb.queues.
  4. Use badmin ckconfig to check the new queue definition. If any errors are reported, fix the problem and check the configuration again.
  5. Run badmin reconfig to reconfigure mbatchd.