My run aspect did ... nothing ? out and err file completely empty

ASPECT
1

Hey,

I’d run two job with ASPECT lately, for simulating convection in the mantle with Rayleigh=1e8, with a walltime of one hour.
When I came back to look at my jobs and how far it gone, by looking the stdout and stderr , this two files was empty, just saying they will create the output directory.
So I looked in it, and there was just the log.txt file, nothing else, even if I told it, in the parameter file, to make checkpoint every 100s.

I have no idea of what happend, or how to fix it. Is it just a bug and I can reload it ?

You have here a google link with the parameter file and the out and err file of the last job ( the ones corresponding to the parameter file): Ra_1e8 - Google Drive
To open the out and err, go on a linux terminal, and tape ‘vi OAR.—name-file’ (replace —name file by the actuel name file).

Thank you,

Emilie.

2

Dear Emilie,

The error message is not very informative, only saying that the job was killed. Can you run ASPECT in debug mode, rather than optimized mode? That should provide a more informative error.

My only other comment is that your input initial temperature file has a surprising name: “output_file_ascii_Ra_1e8.txt”. If you have a lengthy workflow running through several different programs, it might be worth specifying which software created the file in the name so that you don’t get confused later.

3

Emilie:
how big is that input file you’re trying to read? This is the sort of error
you get when you run out of memory.

Best
W.

4

Sorry for the delai, but here I am with news.
I launch aspect in debug mode, but it doesn’t change it, I don’t have more informations.
The parameters.prm file makes 6.9 ko.

5

Dear Emilie,

@bangerth meant how big is your output_file_ascii_Ra_1e8.txt file, not your parameters.prm file.

How big is your output_file_ascii_Ra_1e8.txt file?

Best wishes,
Bob

6

Hi @Emilie,

In parallel to determining the size of output_file_ascii_Ra_1e8.txt, can you also let us know if running the model with the initial temperature prescribed via a function enables the same model to run successfully?

That would also help pinpoint if the issue is indeed related to the initial temperature file.

Snippet below was taken from the linked prm file:

subsection Initial temperature model
  set List of model names = function
  subsection Function
    set Variable names      = x,z
    set Function constants  = p=0.01, L=1, pi=3.1415926536, k=1
    set Function expression = (1.0-z) - p*cos(k*pi*x/L)*sin(pi*z)
  end
end

Cheers,
John

7

Hello !

So, the file for the temperature is 2.6 Mo.
I also launch the job with the function instead of the temperature file for initial temperature condition, but that still doesn’t work, my job run one hour and then, the OAR.—. stderr say that my job was killed (even if it run the entire hour I set up). It doesn’t say why.

Thank you for your help,
Emilie.

8

Hi Emilie,

Thank for the update and it sounds like there is a more fundamental issue going on.

Some follow-up questions to diagnose the issue:

  1. Does stderr provide any additional information about what is happening (is ASPECT launched, etc), or is the file blank?
  2. Are you successfully able to run other models, or is this issue indeed of the PRM file? For example, could you try running one of the cookbook models?
  3. If the above also produces the same issue, can you manually try running ASPECT in the terminal (i.e., not a job submission) by navigating to the folder where ASPECT is installed and running ./aspect? If that produces the normal ASPECT message when no PRM is provided, can you then try running a short cookbook on one processor in the same manner?

Cheers,
John