LSF Version 7.3 - Administering Platform LSF

Administering Platform LSF 515
Running Parallel Jobs
Limitations ◆ A job does not have an estimated start time immediately after mbatchd is
reconfigured.
Backfilling and job slot limits
A backfill job borrows a job slot that is already taken by another job. The backfill
job does not run at the same time as the job that reserved the job slot first.
Backfilling can take place even if the job slot limits for a host or processor have been
reached. Backfilling cannot take place if the job slot limits for users or queues have
been reached.
Configuring backfill scheduling
Backfill scheduling is enabled at the queue level. Only jobs in a backfill queue can
backfill reserved job slots. If the backfill queue also allows processor reservation,
then backfilling can occur among jobs within the same queue.
Configure a backfill queue
1 To configure a backfill queue, define BACKFILL in lsb.queues.
2 Specify Y to enable backfilling. To disable backfilling, specify N or blank space.
Example BACKFILL=Y
Enforcing run limits
Backfill scheduling requires all jobs to specify a duration. If you specify a run time
limit using the command line
bsub -W option or by defining the RUNLIMIT
parameter in
lsb.queues or lsb.applications, LSF uses that value as a hard limit
and terminates jobs that exceed the specified duration. Alternatively, you can
specify an estimated duration by defining the RUNTIME parameter in
lsb.applications. LSF uses the RUNTIME estimate for scheduling purposes
only, and does not terminate jobs that exceed the RUNTIME duration.
Backfill scheduling works most efficiently when all the jobs in a cluster have a run
limit specified at the job level (
bsub -W). You can use the external submission
executable,
esub, to make sure that all users specify a job-level run limit.
Otherwise, you can specify ceiling and default run limits at the queue level
(RUNLIMIT in
lsb.queues) or application level (RUNLIMIT in
lsb.applications).