Terra:Batch Job Files
While not the only method of submitted a job, job files fulfill the needs of most users.
The general idea behind job files follows:
- Make resource requests
- Add your commands and/or scripting
- Submit the job to the batch system
Several of the most important options are described below. These options are typically all that is needed to run a job on Terra.
|Node Count||-N [min[-max]]||-N 4||Spread all cores across 4 nodes|
|Cores Per Node||-ntasks-per-node=|
|Compute Nodes:||256 compute nodes, each with 64GB RAM |
48 GPU nodes, each with a single Tesla K80 GPU and 128GB of RAM
|Interconnect:||Intel OmniPath100 Series switches.|
|Peak Performance:||~X TFLOPs (TBD)|
|Global Disk:||1.5PB (raw) via IBM's GSS26 appliance for general use |
1.5PB (raw) via IBM's GSS256 purchased by a dedicated for GeoPhysics
|File System:||General Parallel File System (GPFS)|
|Batch Facility:||Slurm by SchedMD|
|Location:||Teague Data Center|
|Production Date:||Fall 2016 (tentative)|
It should be noted that Slurm divides processing resources as such: Nodes -> Cores/CPUs -> Tasks
A user may change the number of tasks per core. For the purposes of this guide, all cores will be associated with exactly a single task.
Several examples of Slurm job files for Terra are listed below. For translating Ada/LSF job files, the Batch Job Translation Guide provides some reference.
Documentation for advanced options can be found under Advanced Documentation.