LSF Version 7.3 - Platform LSF Configuration Reference

Description
UNIX only.
Specifies the time interval in seconds between sending SIGINT, SIGTERM, and SIGKILL when
terminating a job. When a job is terminated, the job is sent SIGINT, SIGTERM, and SIGKILL
in sequence with a sleep time of JOB_TERMINATE_INTERVAL between sending the signals.
This allows the job to clean up if necessary.
Default
10
LSB_STOP_ASKING_LICENSES_TO_LS
Syntax
LSB_STOP_ASKING_LICENSES_TO_LS=n | N | y | Y
Description
Used in conjunction with LSF License Scheduler, optimizes the licenses available to pending
jobs. Setting this parameter to y stops mbatchd from requesting a license from the License
Scheduler daemon when a job cannot run—even with a license—because of other pending
reasons. For example, if jobA pends because of a license shortage, and, while jobA waits for a
license, other jobs fill all available slots, mbatchd stops requesting a license for jobA. This
prevents jobA from obtaining a license that it cannot use, which makes more licenses available
for other jobs. Once the slots required by jobA become available, mbatchd requests a license
for jobA. If no license is available, jobA pends until it gets a license.
When set to n (the default), mbatchd requests a license for all pending jobs that require a
license, regardless of additional pending reasons. This ensures that jobs obtain licenses as soon
as possible; jobs can run the moment that all other pending reasons clear. With this parameter
set to n, however, some jobs could pend due to a license shortage. The default setting works
best when jobs in your cluster are unlikely to pend due to resource availability or user limits.
Default
N. A job that was initially pending due to a license requirement and is now pending for another
reason continues to request a license.
LSB_SYNC_HOST_STAT_LIM
Syntax
LSB_SYNC_HOST_STAT_LIM=y | Y
Description
Improves the speed with which mbatchd obtains host status, and therefore the speed with
which LSF reschedules rerunnable jobs: the sooner LSF knows that a host has become
unavailable, the sooner LSF reschedules any rerunnable jobs executing on that host. Useful
for a large cluster.
lsb.params
Platform LSF Configuration Reference 255