USP Zoology HPC Facility Info

Computational resources for the duration of this workshop have been generously provided by the Zoology HPC facility, with special thanks to Diogo Melo for technical support and Roberta Damasceno for coordinating access. The cluster we will be using is located at:

lem.ib.usp.br

The table below indicates all the queues on the Zoology cluster. Queues are partitions of resources that provide tiers of service to users. For example, the ‘short’ queue has limited walltime, but can run up to 10000 jobs per user, whereas the ‘long’ queue can run jobs for a much longer time, but far fewer of them. Values for the various resources in the table below indicate defaults with max values in parentheses. We will primarily be using the ‘proto’ (i.e. prototyping) queue, which will allow us to have interactive access to compute resources (i.e. we will have a command line interface on a compute node). The other queues can be used for longer running jobs, but these must be submitted through a batch job script (which we will cover later in the course given time and interest):

queue	proto	short	long	bigmem
type	interactive	batch	batch	batch
priority	1000	100	100	25
max queuable per user	10	10000	100	10
memory (max)	4 (64)	4 (64)	4 (64)	32 (1280)
n procs (max)	1 (2)	1 (16)	1 (32)	1 (16)
walltime hrs (max)	4 (24)	4 (24)	24 (720)	24 (720)

Example Job Submission Script

Open a new file called ipyrad-anolis.job and add the following text:

#!/bin/bash
#PBS -N ipyrad-anolis
#PBS -l nodes=1:ppn=16
#PBS -l mem=64gb
#PBS -l walltime=2:00:00
#PBS -q short
#PBS -j oe
#PBS -o /home/<username>/ipyrad-workshop/anolis-qsub-job.out

cd /home/<username>/ipyrad-workshop
date > runtime.txt
ipyrad -p params-anolis.txt -s 1234567 -c 16 -f
date >> runtime.txt

Submit this job to the cluster with qsub:

qsub -V ipyrad-anolis.job

And now monitor the progress of the job with qstat:

qstat