Re: [AMBER] Restart the crashed MD run in Amber18

From: Amir Zeb <>
Date: Fri, 18 Jun 2021 10:04:39 +0900

Dear David,

Thanks for the response.
Well, the snapshot of the md info is attached herewith to let you know the
number of steps completed and remaining, when the job is stopped.
I don't know exactly how long it runs and then stops but I'll let you know
my system setup and how it stops.
I'm a terminal user of our server, my machine is connected to the server
via the mobaxterm.
When I'm leaving the office, I'm locking my machine by pressing Wind+L keys.
Since, the simulation job which I have submitted from my machine to the
server via the terminal is running as the background job but stops after
I think that this sleeping mode of the machine may cause the crash of the
simulation job on the server. I'm thinking that if I could put my machine
on sleep mode as never, the simulation running in the background may not

[image: image.png]


On Wed, Jun 16, 2021 at 10:01 PM David A Case <>

> On Wed, Jun 16, 2021, Amir Zeb wrote:
> >
> >in the restrated simulation. The job restarted
> >successfully but stopped/crashed again after some time. Would you please
> >let me know how I can find the reason for this crash?
> So, now it seems that the instability is reproducible. How long does it
> take after the restart before you encounter problems? It's often useful
> to run shorter simulations with small values of ntpr to get more detailed
> information.
> One place to start is to use the "checkstructure" action in cpptraj on the
> restart coordinates. That might flag problems.
> > nstlim=200000000,
> > ntpr=10000,
> Consider *much* shorter runs, setting ntpr to 10 or 100. Your goal is to
> collect some information near the point at which the crash happens.
> ....dac
> _______________________________________________
> AMBER mailing list

AMBER mailing list

(image/png attachment: image.png)

Received on Thu Jun 17 2021 - 18:30:02 PDT
Custom Search