open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,
I have a problem with my simulation over a year. It starts very well, after 8 hours of time the log file stops writing to its file, until the end of the time given for the simulation.
And I don't get any error message. I don't know if this is related to the memory of my server used for the calculation or others.
You will find attached an example of my log file.
Thank you for your help.

We cannot acces to the file.
Change extension to txt!

Hello again,
Attached the log file in .txt

Hi
Nothings particular to say in this file...
As it's a parallel run, it's hard to say with only 1 log file.
Try to look if all other doesn't have any other information...
You could also try to use the last available time step as an initial condition and see if the simulation goes further or not...
This will help you to identify if there is a problem in the model at this moment or if it's a server problem...
We observed in some case a kind of memory leak in parallel which increase the memory and leads to a crash when memory is full. This could be seen by following the execution process on nodes...

Regards

Hello Yann,

As told in other topics, if you look at the PE*LOG files, the subdomain which may be suspicious and where the error may occur with extra error message is the one (or the ones) which has/have a different size.

Moreover, with my OS, I have an error message for the master proc, like:

application called MPI_Abort(MPI_COMM_WORLD, 2) - process 32
Traceback (most recent call last):

and it gives me the number of the subdomain where an error may occur: 32 in that case.

Hope this helps,

Chi-Tuan