Scripts to manage NVIDIA GPU devices in SGE 6.2u5.
The last Sun Grid Engine that is packaged in Ubuntu 14.04 LTS does not contain the RSMAP functionality that is implemented in recent Univa Grid Engine. The ad-hoc scripts in this package implement resource allocation for NVIDIA devices.
First, set up consumable complex gpu
.
qconf -mc
#name shortcut type relop requestable consumable default urgency
#----------------------------------------------------------------------------------------------
gpu gpu INT <= YES JOB 0 0
At each exec-host, add gpu
resource complex. For example,
qconf -aattr exechost complex_values gpu=1 node01
Set up prolog
and epilog
in the queue.
qconf -mq gpu.q
prolog sgeadmin@/path/to/sge-gpuprolog/prolog.sh
epilog sgeadmin@/path/to/sge-gpuprolog/epilog.sh
Alternatively, you may set up a parallel environment for GPU and set
start_proc_args
and stop_proc_args
to the packaged scripts.
Request gpu
resource in the designated queue.
qsub -q gpu.q -l gpu=1 gpujob.sh
The job script can access CUDA_VISIBLE_DEVICES
variable.
#!/bin/sh
echo $CUDA_VISIBLE_DEVICES
The variable contains a comma-delimited device IDs, such as 0
or 0,1,2
depending on the number of gpu
resources to be requested. Use the device ID
for cudaSetDevice()
.