Ciao Yvan,
I still don't have any answer for sending you the files...
I've tried what you wrote me in the previous post, but it doesen't work. Maybe in some other subroutine, it's written to write the files in another format and Saturne takes in consideration the other subroutine first? But I don't find in any of them any command for output files. The whole case I was given is without interface and was not using this cs_user_postprocess.c subroutine, so how could Saturne knows how to write the files? there's a default value when nothing is precised?
I send you the cs_user_postprocess.c as I modified it...
Best regards
Christine
PS: about the RESTART, I should have this ILEAUX and IECAUX files?
____________________
Attachment:
cs_user_postprocess.c [22.2 KiB]
RESTART
Forum rules
Please read the forum usage recommendations before posting.
Please read the forum usage recommendations before posting.
-
Yvan Fournier
Re: RESTART
Hello,
Line 138 of your cs_user_postprocess.c, replace "false" by "true" to activate the option.
Regarding ILEAUX and IECAUX, these options indicate that auxiliary restart files (face values, with mass fluxes) will be used if a main restart file is used. By default, we always leave them at 1, but if you run with IECAUX = 0, you will only have a have a main restart file, and will need to read using ILEAUX = 0 also.
Best regards,
Yvan
Line 138 of your cs_user_postprocess.c, replace "false" by "true" to activate the option.
Regarding ILEAUX and IECAUX, these options indicate that auxiliary restart files (face values, with mass fluxes) will be used if a main restart file is used. By default, we always leave them at 1, but if you run with IECAUX = 0, you will only have a have a main restart file, and will need to read using ILEAUX = 0 also.
Best regards,
Yvan
-
Christine Leroy
Re: RESTART
Previously Yvan Fournier wrote:
Thanks for your reply, I replace and it's ok now, I have the files in the text format.
About the restart, it's 'better' now but still not ok:
- The options were already at 1, and I let like this, but I don't have any Ileaux or Iecaux files, I don't understand...
- There was a problem with the cs_user_scripts.py file : at line 61, it's written "domain.restart_input= ...", but reading above (line 57), I saw that it should have been "domain.restart_input_dir=...". So I changed, and it's ok, I could restart from previous run without any error message. But instead of going on with new iterations, i's running again from beginning...so no interest !!!
PS. everytime saturne is running for more than few iterations, I have a "kernel crash"...saturne is still going on but I loose the internet connection...have you ever heard of this kind of problem?
Best regards
Christine
ciao Yvan,Hello,
Line 138 of your cs_user_postprocess.c, replace "false" by "true" to activate the option.
Regarding ILEAUX and IECAUX, these options indicate that auxiliary restart files (face values, with mass fluxes) will be used if a main restart file is used. By default, we always leave them at 1, but if you run with IECAUX = 0, you will only have a have a main restart file, and will need to read using ILEAUX = 0 also.
Best regards,
Yvan
Thanks for your reply, I replace and it's ok now, I have the files in the text format.
About the restart, it's 'better' now but still not ok:
- The options were already at 1, and I let like this, but I don't have any Ileaux or Iecaux files, I don't understand...
- There was a problem with the cs_user_scripts.py file : at line 61, it's written "domain.restart_input= ...", but reading above (line 57), I saw that it should have been "domain.restart_input_dir=...". So I changed, and it's ok, I could restart from previous run without any error message. But instead of going on with new iterations, i's running again from beginning...so no interest !!!
PS. everytime saturne is running for more than few iterations, I have a "kernel crash"...saturne is still going on but I loose the internet connection...have you ever heard of this kind of problem?
Best regards
Christine
-
Jacques Fontaine
Re: RESTART
Hello,
The comment domain.restart_input_dir is a mistake (corrected in 2.2). You just need to complete domain.restart_input='RESU/<run_id>/checkpoint' (do not forget the single quotes 'x').
About the kernel crash, I have never seen this problem before.
Best regards,
JF
The comment domain.restart_input_dir is a mistake (corrected in 2.2). You just need to complete domain.restart_input='RESU/<run_id>/checkpoint' (do not forget the single quotes 'x').
About the kernel crash, I have never seen this problem before.
Best regards,
JF
-
Christine Leroy
Re: RESTART
ok so back to the same error as before (SIGSEGV...)
what do you mean by "single quotes 'x' "? I'm not using any interface.
I was sent 3 files from which I could do a restart : suiava, suiavx and rayava. But, same problem of course if I try the restart.
My question is: where do these files come from? is it from another version of saturne? I'm working with the 2.1.4 version....when I run a simulation, no files of this kind are generated...just "auxiliary", "main" and "radiative" transfer in the checkpoint directory. Is it normal?
sorry for all theses questions, but I really need to solve this problem of restart.
Thanks again
Christine
what do you mean by "single quotes 'x' "? I'm not using any interface.
I was sent 3 files from which I could do a restart : suiava, suiavx and rayava. But, same problem of course if I try the restart.
My question is: where do these files come from? is it from another version of saturne? I'm working with the 2.1.4 version....when I run a simulation, no files of this kind are generated...just "auxiliary", "main" and "radiative" transfer in the checkpoint directory. Is it normal?
sorry for all theses questions, but I really need to solve this problem of restart.
Thanks again
Christine
-
Jacques Fontaine
Re: RESTART
These single quotes: 'RESU/<run_id>/checkpoint '
suiava, suiavax, rayava are restart files from the 1.3/2.0 version of Code_Saturne. They are compatibles with the 2.1 version but you need to rename suiava into main, suiavx into axiliary and rayava into radiative.
Best regards,
JF
suiava, suiavax, rayava are restart files from the 1.3/2.0 version of Code_Saturne. They are compatibles with the 2.1 version but you need to rename suiava into main, suiavx into axiliary and rayava into radiative.
Best regards,
JF
-
Christine Leroy
Re: RESTART
Hello Jacques,
Yes the person who sent me the files explained me this, but even changing the names, the problem remains...and the quotes were already ok.
I've send all the case at the address Yvan gave me (saturne-support@edf.fr) if it can help finding the problem
Tank you very much again and sorry about all this confusion about the files...
Christine
Yes the person who sent me the files explained me this, but even changing the names, the problem remains...and the quotes were already ok.
I've send all the case at the address Yvan gave me (saturne-support@edf.fr) if it can help finding the problem
Tank you very much again and sorry about all this confusion about the files...
Christine
-
Yvan Fournier
Re: RESTART
Hello,
I just tried running your case, and I do not have a SIGSEGV, but I do have another error (I ran 3 iterations, renamed checkpoint to restart, and restarted with that, so the situation might be slightly different than using files generated with version 2.0).
The first error (as determined by running under Valgrind) occurs in src/comb/cs_coal_physprop.f90, line 401, due to izone not being initialized (meaning that iffppp(ifac) is not initialized, for face 1).
I am not familiar with the coal combustion version, so I do not know if this is a user subroutine error, or if this is a bug in the code, so at this stage, I'll check with colleagues, and keep you informed.
Best regards,
Yvan
I just tried running your case, and I do not have a SIGSEGV, but I do have another error (I ran 3 iterations, renamed checkpoint to restart, and restarted with that, so the situation might be slightly different than using files generated with version 2.0).
The first error (as determined by running under Valgrind) occurs in src/comb/cs_coal_physprop.f90, line 401, due to izone not being initialized (meaning that iffppp(ifac) is not initialized, for face 1).
I am not familiar with the coal combustion version, so I do not know if this is a user subroutine error, or if this is a bug in the code, so at this stage, I'll check with colleagues, and keep you informed.
Best regards,
Yvan
-
Christine Leroy
Re: RESTART
Ciao Yvan,
any news about the case I sent you for which I can't do any restart?
Best Regards
Christine
any news about the case I sent you for which I can't do any restart?
Best Regards
Christine
-
Yvan Fournier
Re: RESTART
Hello Christine,
Could you please switch to the new Code_Saturne forum at http://www.code-saturne.org/forum ?
I do not have any news about the restart functionality for the new combustion model. The combustion team (who I believe are the ones who provided you the test case, and who developed the model in the first place) is now aware f the bug (or so I have been told), and should in theory prepare the missing development and a test case, so that the core team can integrate it, but I am not sure they have started yet. I'll try to get an update on the status of this bug, and remind them that you need it.
Best regards,
Yvan
Could you please switch to the new Code_Saturne forum at http://www.code-saturne.org/forum ?
I do not have any news about the restart functionality for the new combustion model. The combustion team (who I believe are the ones who provided you the test case, and who developed the model in the first place) is now aware f the bug (or so I have been told), and should in theory prepare the missing development and a test case, so that the core team can integrate it, but I am not sure they have started yet. I'll try to get an update on the status of this bug, and remind them that you need it.
Best regards,
Yvan