[Mesa-users] Error in make_co_wd test case (r24.08.1)

[Farag, Ebraheem]

Hello Mike,

Your error message seems to indicate you've placed the namelist options into the &controls section of your inlist instead of the &star_job section. Please see https://docs.mesastar.org/en/24.08.1/reference/star_job.html#eosdt-cache-dir.

-EbF
________________________________
From: Huaiyu Duan <duan at unm.edu>
Sent: Friday, September 12, 2025 5:16 PM
To: Farag, Ebraheem <ebraheem.farag at yale.edu>
Cc: mesa-users at lists.mesastar.org <mesa-users at lists.mesastar.org>
Subject: Re: Error in make_co_wd test case (r24.08.1)

Hi Ebraheem,

Thanks for the suggestion! But the code immediately failed with the following error message:
============================
At line 671 of file ../private/ctrls_io.f90
Fortran runtime error: Cannot match namelist object name eosdt_cache_dir
============================

— Huaiyu (Mike)

On Sep 12, 2025, at 10:17 AM, Farag, Ebraheem <ebraheem.farag at yale.edu> wrote:

  [EXTERNAL]
Hello Mike,

Yes, MESA creates caches in the MESA directory during runs for various things. If you need the caches to be local due to file write permissions, try adding these options to your star_job section of the inlists and it might resolve this conflict.

! local cache directories data
      eosDT_cache_dir      = './eosDT_cache'
      kap_cache_dir        = './kap_cache'
      rates_cache_dir      = './rates_cache'

-EbF
________________________________
From: Mesa-users <mesa-users-bounces at lists.mesastar.org> on behalf of Huaiyu Duan via Mesa-users <mesa-users at lists.mesastar.org>
Sent: Friday, September 12, 2025 12:05 PM
To: mesa-users at lists.mesastar.org <mesa-users at lists.mesastar.org>
Subject: [Mesa-users] Error in make_co_wd test case (r24.08.1)

Hello,

I tried to run the make_co_wd test case in r24.08.1 but failed. I’m very new to MESA, and I don’t completely understand the error messages it generated. (See below for the last few lines of the output.) Here is my guess, though.

I had MESA set up as a systemwide package so that the students in my class can avoid all the technical issues of installation. I copied make_co_wd from the test_suite directory to my own scratch space and ran it there. But somehow MESA wanted to copy some of the stuff generated in during the run back to the main MESA directory which is forbidden.

Is my interpretation correct? If so, is there way to disable this behavior or is MESA not designed to be shared by many users who may run it simultaneously? Thank you!

================================
 retry:  adjust_correction failed in eval_equations -- give up in solver     322
write /opt/local/mesastar/24.08.1/data/kap_data/cache/gs98_co_z0.02_x0.0.bin
 mv failed for './.mesa_temp_cache/kap_cache/gs98_co_z0.02_x0.0.bin' '/opt/local/mesastar/24.08.1/data/kap_data/cache/gs98_co_z0.02_x0.0.bin' skipping
 hydro_solver_step returned ierr          -1
 s% model_number         322
 nz        1026
 s% num_retries          51

 retry:  adjust_correction failed in eval_equations -- give up in solver     322
write /opt/local/mesastar/24.08.1/data/kap_data/cache/gs98_co_z0.02_x0.0.bin
 mv failed for './.mesa_temp_cache/kap_cache/gs98_co_z0.02_x0.0.bin' '/opt/local/mesastar/24.08.1/data/kap_data/cache/gs98_co_z0.02_x0.0.bin' skipping
terminated evolution: nonzero_ierr

                 runtime (minutes), retries, steps       11.52        52       322
DATE: 2025-09-11
TIME: 21:09:49

******************************************************************
failed to create end_he_core_burn.mod when running inlist_to_end_he_core_burn_header
******************************************************************
================================

— Huaiyu (Mike)

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20250912/76af8148/attachment.htm>