[gmx-users] Segmentation fault mdrun

[Mark Abraham]

Hi,

For example, Martini's webpages don't claim 40fs is universally applicable,
so I'd be thinking in that direction... Perhaps you can increase the time
step after you equilibrate, but just as with atomistic simulations, you
might need to be more gentle while you are equilibrating. Also, your tau-t
is on the short side for that time step.

Mark


On Mon, Sep 1, 2014 at 5:26 PM, Alex s <as1783 at hotmail.com> wrote:

>
>
>
> They're ions that I've just added to the water phase of the system. That
> was a case I highlighted to show how sensitive the system is to minor
> changes. I have other systems which are ion free and still experience
> segmentation fault.
> I've just reviewed all my md runs for my systems, I have found that 2 of
> the systems which experienced segmentation fault had a warning that the
> pressure scaling was more than 1 %  . However the other systems that
> experienced segmentation fault presented no such warning yet the force
> imbalance was as high as 6 %. Could this be the issue?
> Thanks
> Alex
>
> > Date: Sun, 31 Aug 2014 17:14:48 -0400
> > From: jalemkul at vt.edu
> > To: gmx-users at gromacs.org
> > Subject: Re: [gmx-users] Segmentation fault mdrun
> >
> >
> >
> > On 8/31/14, 1:55 PM, Alex s wrote:
> > >
> > > There was a typo in the last message, I found NO issues in point 2.
> >
> > Actually, you did.  You said when you removed the ions, the problem was
> > resolved.  That indicates to me that the ions are somehow unstable.  I
> don't
> > know much about doing CG simulations, so maybe someone else can advise,
> but it
> > seems that the long time step is causing large forces to accumulate on
> the ions
> > and you get the crash.  Are the ion particles just a single ion or some
> > representation of a hydrated ion?
> >
> > -Justin
> >
> > > From: as1783 at hotmail.com
> > > To: gmx-users at gromacs.org
> > > Date: Sun, 31 Aug 2014 18:52:39 +0100
> > > Subject: Re: [gmx-users] Segmentation fault mdrun
> > >
> > > Thanks for the reply
> > > I've attached a copy of my mdp file. It is a coarse grained system so
> a time step of 0.04 shouldn't be a problem. Regarding the link you've
> given, points 1 and 4 are not applicable to my situation, I've tried point
> 2 and found issues. What should I be looking for in point 3?
> > > Thanks
> > > Alex
> > >> Date: Sun, 31 Aug 2014 13:20:31 -0400
> > >> From: jalemkul at vt.edu
> > >> To: gmx-users at gromacs.org
> > >> Subject: Re: [gmx-users] Segmentation fault mdrun
> > >>
> > >>
> > >>
> > >> On 8/31/14, 1:17 PM, Alex s wrote:
> > >>> Hi
> > >>> I'd really appreciate some advice on this issue, I tend to get a
> segmentation fault during my mdrun. I've searched online for ways to
> resolve this and tried to diagnose whats causing this problem but I'm
> having no luck.
> > >>> I know that if you get a segmentation fault at step 0 its most
> likely a problem with system configuration or a minimization issue. However
> for my system I successfully run energy minimization and equilibration.
> When I run md the segmentation fault occurs usually very late on the system
> (past the 100ns mark).
> > >>> I've found that minor adjustments to the system can either lead to
> the error or remove it. For example for one system, I ran it ION free and I
> had no issues, however when I just replace 12 water molecules with NA+ and
> CL- this leads to a segmentation fault. Also I was able to fix this error
> in another system just by changing the time step from to 0.04 to 0.03.
> > >>>   From what I'm describing, does it seem as if the fault lies with
> the system set up? or the mdp file? or could it be a compiler issue? What
> information should I be looking for in my output file to perhaps determine
> where issue lays?
> > >>
> > >> Provide the full .mdp file and people can comment on it.  Is this an
> atomistic
> > >> system?  A time step of 0.04 is insane if it is.
> > >>
> > >> Diagnostics:
> > >>
> http://www.gromacs.org/Documentation/Terminology/Blowing_Up#Diagnosing_an_Unstable_System
> > >>
> > >> -Justin
> > >>
> > >> --
> > >> ==================================================
> > >>
> > >> Justin A. Lemkul, Ph.D.
> > >> Ruth L. Kirschstein NRSA Postdoctoral Fellow
> > >>
> > >> Department of Pharmaceutical Sciences
> > >> School of Pharmacy
> > >> Health Sciences Facility II, Room 601
> > >> University of Maryland, Baltimore
> > >> 20 Penn St.
> > >> Baltimore, MD 21201
> > >>
> > >> jalemkul at outerbanks.umaryland.edu | (410) 706-7441
> > >> http://mackerell.umaryland.edu/~jalemkul
> > >>
> > >> ==================================================
> > >> --
> > >> Gromacs Users mailing list
> > >>
> > >> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> posting!
> > >>
> > >> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> > >>
> > >> * For (un)subscribe requests visit
> > >> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users
> or send a mail to gmx-users-request at gromacs.org.
> > >
> > >
> >
> > --
> > ==================================================
> >
> > Justin A. Lemkul, Ph.D.
> > Ruth L. Kirschstein NRSA Postdoctoral Fellow
> >
> > Department of Pharmaceutical Sciences
> > School of Pharmacy
> > Health Sciences Facility II, Room 601
> > University of Maryland, Baltimore
> > 20 Penn St.
> > Baltimore, MD 21201
> >
> > jalemkul at outerbanks.umaryland.edu | (410) 706-7441
> > http://mackerell.umaryland.edu/~jalemkul
> >
> > ==================================================
> > --
> > Gromacs Users mailing list
> >
> > * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> posting!
> >
> > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> >
> > * For (un)subscribe requests visit
> > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-request at gromacs.org.
>
>
> --
> Gromacs Users mailing list
>
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-request at gromacs.org.
>