LSB_CGROUP_CPU_SHARES_OLD_DEFAULT
On hosts enabled for v1 or v2 Linux cgroups and
configured with LSF_LINUX_CGROUP_ACCT=Y in the lsf.conf
file, this parameter controls the default or initial value of the cpu.shares and
cpu.weight cgroup interface values that LSF should
consider when creating the cgroup for the job. Prior to Fix Pack 15, LSF used
the Linux OS default for cpu.shares and cpu.weight values.
Starting in Fix Pack 15, by default (that is, if
LSB_CGROUP_CPU_SHARES_OLD_DEFAULT is set to N|n or is
not defined), cpu.shares and cpu.weight values are relative to the
number of tasks the job will run on a particular execution host.
Syntax
LSB_CGROUP_CPU_SHARES_OLD_DEFAULT=Y|y|N|n
Description
When set to N|n (or undefined), the initial values for
cpu.shares and cpu.weight values will be relative to the number of
tasks the job will run on a particular execution host; this way, jobs that are less CPU intensive
can then be given a fraction of the CPU share or weight of other jobs by also setting in the
CGROUP_CPU_SHARES_FACTOR parameter in the lsb.applications and lsb.queues files (for example,
CGROUP_CPU_SHARES_FACTOR=25), rather than receive the same share or weight of
the CPU as all other running jobs. The initial values for cpu.shares and
cpu.weight will be scaled by this CGROUP_CPU_SHARES_FACTOR
value if configured.
Default
N