[gmx-users] Help with interpreting message for fatal error

Thu Mar 15 11:19:38 CET 2018

Hi Justin,

Thanks again for the tip. I'm always assuming that there's something wrong
with my setup but I tested using one rank only and then I get an order of
magnitude smaller fluctuations... #underInvestigation

--
Viveca

On Mon, Mar 12, 2018 at 6:42 PM, Justin Lemkul <jalemkul at vt.edu> wrote:

>
>
> On 3/12/18 1:40 PM, Viveca Lindahl wrote:
>
>> Hi Justin,
>>
>> I'm using semi-isotropic pressure coupling and the periodicity is along
>> z. I
>> should say that I'm also adaptively biasing a base pair to open, which is
>> triggering the fluctuations.
>>
>> I guess you mean using only OpenMP? I have not tried that no, thanks for
>> the suggestion. Now I changed to Berendsen coupling which avoids the
>> problem (at least the fatal errors).
>>
>
> Yes, the problem is coming from either DD bookkeeping or actual
> instability. Trying with only OpenMP parallelization and therefore no
> domains will sort out whether it's physical or just a software problem.
> There's a lot going on if you've got periodic molecules, NPT, and a biasing
> potential, all with DD. Probably something no one has ever tested in
> GROMACS.
>
>
> -Justin
>
> --
>> Viveca
>>
>>
>> On Mon, Mar 12, 2018 at 3:46 PM, Justin Lemkul <jalemkul at vt.edu> wrote:
>>
>>
>>> On 3/12/18 10:42 AM, Viveca Lindahl wrote:
>>>
>>> Hi Gregory,
>>>>
>>>> In the sense that I see it happening repeatedly yes :) But I can't
>>>> reproduce it very predictably.
>>>>
>>>> After seeing that (for my small system) I get very large box vector
>>>> fluctuations using Parrinello-Rahman, my current assumption is that
>>>> there
>>>> are some stability assumptions made by mdrun internally that are being
>>>> violated.
>>>> I would still be interested in understanding what actually goes wrong
>>>> though.
>>>>
>>>> Periodic molecules are usually used in CNT systems or similar, and the
>>> typical advice is to not apply pressure coupling along the long axis. I
>>> suspect interactions get lost when the box fluctuates and the domain
>>> topology does not get updated correctly. Does it work correctly with
>>> OpenMP?
>>>
>>> -Justin
>>>
>>>
>>> --
>>>
>>>> Viveca
>>>>
>>>>
>>>> On Mon, Mar 12, 2018 at 4:49 PM, Gregory Poon <gpoon at gsu.edu> wrote:
>>>>
>>>> Since there hasn't been any response to your question yet, may I ask: is
>>>>
>>>>> it reproducible?
>>>>>
>>>>> Gregory
>>>>>
>>>>>
>>>>> On 3/11/2018 7:30 AM, Viveca Lindahl wrote:
>>>>>
>>>>> Hi,
>>>>>
>>>>>> I'm running a simulation of a periodic DNA molecule. After some time I
>>>>>> get,
>>>>>> sometimes the following error message:
>>>>>>
>>>>>> ---------------
>>>>>>
>>>>>> A list of missing interactions:
>>>>>>             Proper Dih. of   3917 missing      1
>>>>>>
>>>>>> Molecule type 'DNA_chain_A'
>>>>>> the first 10 missing interactions, except for exclusions:
>>>>>>             Proper Dih. atoms  414  417  418  421 global   414   417
>>>>>>   418
>>>>>> 421
>>>>>>
>>>>>> -------------------------------------------------------
>>>>>> Program:     gmx mdrun, version 2018.1-dev-20180306-33093601f
>>>>>> Source file: src/gromacs/domdec/domdec_topology.cpp (line 436)
>>>>>> MPI rank:    0 (out of 32)
>>>>>>
>>>>>> Fatal error:
>>>>>> 1 of the 10381 bonded interactions could not be calculated because
>>>>>> some
>>>>>> atoms
>>>>>> involved moved further apart than the multi-body cut-off distance
>>>>>> (0.5465
>>>>>> nm)
>>>>>> or the two-body cut-off distance (1.093 nm), see option -rdd, for
>>>>>> pairs
>>>>>> and
>>>>>> tabulated bonds also see option -ddcheck
>>>>>>
>>>>>> For more information and tips for troubleshooting, please check the
>>>>>> GROMACS
>>>>>> website at https://na01.safelinks.protect
>>>>>> ion.outlook.com/?url=http%3A%
>>>>>> 2F%2Fwww.gromacs.org%2FDocumentation%2FErrors&data=02%7C01%
>>>>>> 7Cgpoon%40gsu.edu%7C18eac9ccdd4b479d86c708d5875cbb61%7C515ad
>>>>>> 73d8d5e4169895c9789dc742a70%7C0%7C0%7C636563754716997653&
>>>>>> sdata=wUdEASAxNIPHNAmQwBO8EO0Ec8EI%2FVdjY7f7i9R0IYc%3D&reserved=0
>>>>>> -------------------------------------------------------
>>>>>> Rank 0 [Sun Mar 11 13:37:32 2018] [c0-0c1s3n0] application called
>>>>>> MPI_Abort(MPI_COMM_WORLD, 1) - process 0
>>>>>> _pmiu_daemon(SIGCHLD): [NID 00076] [c0-0c1s3n0] [Sun Mar 11 13:37:32
>>>>>> 2018]
>>>>>> PE RANK 0 exit signal Aborted
>>>>>> [NID 00076] 2018-03-11 13:37:32 Apid 12709959: initiated application
>>>>>> termination
>>>>>> Application 12709959 exit codes: 134
>>>>>> Application 12709959 resources: utime ~1083s, stime ~16s, Rss ~21376,
>>>>>> inblocks ~34261, outblocks ~55016
>>>>>> ---> Reached end of run script.
>>>>>>
>>>>>>
>>>>>> How should I... proceed here? Is my system set up poorly, is it a bug
>>>>>> in
>>>>>> mdrun? Suggestions appreciated.
>>>>>>
>>>>>> --
>>>>>> Viveca
>>>>>>
>>>>>> --
>>>>>>
>>>>> *Gregory M. K. Poon, PhD, RPh*
>>>>> Associate Professor
>>>>> Departments of Chemistry and Nutrition | Georgia State University
>>>>> NSC 414/415/416 | 50 Decatur St. SE, Atlanta, GA 30302
>>>>> P.O. Box 3965 | Atlanta, GA 30303
>>>>> Ph (404) 413-5491 | gpoon at gsu.edu <mailto:gpoon at gsu.edu>
>>>>>
>>>>> --
>>>>> 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.
>>> Assistant Professor
>>> Virginia Tech Department of Biochemistry
>>>
>>> 303 Engel Hall
>>> 340 West Campus Dr.
>>> Blacksburg, VA 24061
>>>
>>> jalemkul at vt.edu | (540) 231-3129
>>> http://www.biochem.vt.edu/people/faculty/JustinLemkul.html
>>>
>>> ==================================================
>>>
>>>
>>> --
>>> 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.
> Assistant Professor
> Virginia Tech Department of Biochemistry
>
> 303 Engel Hall
> 340 West Campus Dr.
> Blacksburg, VA 24061
>
> jalemkul at vt.edu | (540) 231-3129
> http://www.biochem.vt.edu/people/faculty/JustinLemkul.html
>
> ==================================================
>
> --
> 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.
>