Using the bioxcluster

Some info on the bioxcluster:

To open the firewall, eg #usernum=152, #sunetid=brian
ssh -p 259 user#usernum:skey@bioxcluster.stanford.edu
To ssh in there:
ssh -p 1000+#usernum #sunetid@bioxcluster.stanford.edu
To scp files to the cluster:
scp -oport=1000+#usernum file #sunetid@bioxcluster.stanford.edu:./
Submitting a job:
qsub job.sh
Deleting a job:
qdel job.sh
What jobs are running:
qstat, or showq (slow)

Format of simplest job.sh:
#!/bin/bash
#PBS -l nodes=1,walltime=24:00:00
#PBS -N my_job

export tempdir=$TMPDIR # temp space on local compute node
export inputdir=$HOME/dir1 # location of input files
export outputdir=$HOME/dir2 # location to put results into
export output=outfile # file for redirected standard output
export error=errfile # file for redirected standard error

cp -r $inputdir/* $tempdir

cd $tempdir

python run_program.py 1>$output 2>$error

cp outfile errfile $outputdir