Running Slurm inside Docker

[Feriolet]

Hi! Can I know how can we run sbatch inside the docker? I believe the docker does not have slurm install, and it takes me a very long time to do a masif search for 1 protein-ligand complex.

Thanks!

[arneschneuing]

Hi @Feriolet,
I am not a docker expert but I believe the best solution depends on your system. It is probably possible to use our provided docker as a base image to create a new one in which you install slurm. I personally use apptainer and launch a separate container in each job similar to this example.

[Feriolet]

I see. Do you think it is faster to use slurm instead of regular bash? My experience of using the benchmarking run is quite slow, around 30 min for 1 search, matching around 5-7 proteins. They utilise almost all of my CPU, so I thought that should be the best I could do.

[arneschneuing]

Do I understand correctly that you get 5-7 results from the search? How big is the database you search against?
Running it sequentially is really slow. I usually parallelize large-scale searches across 500-1000 CPUs.

[Feriolet]

It's been a long time, but I think I meant that I am searching for matches from a list of 5-7 proteins. I am using 200 CPU, but I am not sure if I did it sequentially or parallelise it. How can you check if it is run sequentially?