CRAZY crashes without error

Alfio Lazzaro alfio.... at gmail.com
Tue Oct 17 14:25:56 UTC 2017


Dear Matthijs,
I took your input file and run the test on my local machine (9 MPI ranks + 
2 OpenMP threads).
I'm using CP2K trunk version with GCC 5.3.0. 
I can run up to 4000 steps and I don't get any error...

Could you reproduce the error on your cluster? How do you compile the code? 
I would also suggest compiling CP2K with debug symbols. 

Alfio

Il giorno domenica 15 ottobre 2017 21:05:36 UTC+2, Matthijs Lasure ha 
scritto:
>
>  Hi all,
>
> I'm a newcomer to CP2K, and I have a problem with a MD simulation of 
> chloropropanoic acid.
> I'm trying to do a 60.000 steps simulation, very simular to the tutorial 
> simulation with travis. 
> (http://www.travis-analyzer.de/files/travis_ir_raman.pdf, as found under 
> the howto section)
>
> Whilst the example ran fine, this one crashed now three times without an 
> error in the output file.
> The output from the job manager just speaks of a process terminating 
> unexpected (see the .o file). No additional information is provided.
>
> I've also noticed that the calculation time per step has tenfolded shortly 
> before the crash. Could this be related?
>
> Included are the job information files, the time graph, input file and 
> last 500 lines of the output.
>
> Thanks in advance,
> With kind regards,
>
> Matthijs Lasure
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20171017/f9338487/attachment.htm>


More information about the CP2K-user mailing list