Running a job on BEDE
By: Henry Oldroyd
Step 1: Login to BEDE by SSH
Enter the password for your machine account as the first factor and then multi-factor authentication as the second factor
PS C:\Users\henry> ssh login1.bede.dur.ac.uk
(henryoldroyd@login1.bede.dur.ac.uk) First Factor:
(henryoldroyd@login1.bede.dur.ac.uk) Second Factor:
============
Bede - N8CIR
============
For documentation, see:
https://bede-documentation.readthedocs.io/en/latest/
There is a Slack workspace for general Bede questions. To request access:
https://n8cir.org.uk/supporting-research/facilities/bede/rse-support-bede/
Interested in a workshop on the effective use of BEDE? Let us know here:
https://forms.gle/cjNJrqUkMEw9AYUm8
Last login: Wed Sep 4 08:19:18 2024 from 90.248.225.180
[henryoldroyd@login1.bede ~]$
Step 2:
Make a job file based on the SLURM syntax:
[henryoldroyd@login1.bede ~]$ cat job.com
#!/bin/bash
#SBATCH --account=bdlan09
#SBATCH --TIME=10:0
echo "echo job on bede"
[henryoldroyd@login1.bede ~]$
Step 3: Make a login session by providing the project code
ghlogin -A bdlan09
Step 4: Run the job with srun This seems like it should be a simple but I have been having repeated issues with the submitted job not running
[henryoldroyd@login2.bede ~]$ ghlogin -A bdlan09
srun: job 839128 queued and waiting for resources
srun: job 839128 has been allocated resources
[henryoldroyd@gh001.bede ~]$ ls
hello.py job.com
[henryoldroyd@gh001.bede ~]$ srun job.com
srun: Job 839128 step creation temporarily disabled, retrying (Requested nodes are busy)
srun: Job 839128 step creation still disabled, retrying (Requested nodes are busy)
srun: Job 839128 step creation still disabled, retrying (Requested nodes are busy)
srun: Job 839128 step creation still disabled, retrying (Requested nodes are busy)
srun: Job 839128 step creation still disabled, retrying (Requested nodes are busy)
srun: Job 839128 step creation still disabled, retrying (Requested nodes are busy)