From: Giacomo Fiorin (giacomo.fiorin_at_gmail.com)
Date: Tue Apr 30 2024 - 16:09:48 CDT

Hi Jana,

How do I submit “the next job”?
>
>
Handling dependent jobs depends on the type of scheduler that is being used
in the cluster. The most common is Slurm, where you would add a --depend
flag to sbatch: https://urldefense.com/v3/__https://slurm.schedmd.com/sbatch.html__;!!DZ3fjg!4BlyKbCtccdyXWLc0e6NV1wDT4gV3JuXTAmqNv8WYYlUZSrXIisg6yKKm_mdJsW4x2ZX5da97QyaFG36ivgT5EjL2A$ But the specifics
will be decided by your cluster admins: documenting the batch system is
part of their core duties.

Regards
Giacomo

>
>
> Thanks,
>
> Jana
>
>
>
>
>
>
>
>
>
> Giacomo
>
>
>
> On Tue, Apr 30, 2024 at 2:49 PM Sefcikova, Jana <
> j.sefcikova_at_northeastern.edu> wrote:
>
> Hi,
>
> Is it possible to restart calculations (either in Minimization or MD
> simulation step of QwikMD) from the last step? If so, HOW?
>
> I am running QwikMD on a university computer cluster with assigned/limited
> time and processors allocation. The cluster administrators assigned to my
> account 24 hours and 28 processors, which is not enough to finish either
> 500000-steps minimization or 500000-steps MD simulation. After 24 hours, my
> jobs are terminated including VMD.
>
> Thanks,
> Jana
>
>