-
Notifications
You must be signed in to change notification settings - Fork 365
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
case.submit does not display system error messages #1396
Comments
Thanks - that is probably what I remember - Rob volunteering to create this issue on the ESMCI side, based on a Slack conversation. |
Update CIME to cime5.3 Update ACME to use CIME 5.3.x up to ESMCI hash 6156e0a from Thursday 4/13. scripts_regression_tests and acme_developer tests pass on penn workstation. This CIME change involves a large cosmetic change in CIME. Several directories have been renamed. The main interface to CIME, the scripts dir remains unchanged. cime/cime_config is renamed to cime/config The Fortran source code has been consolidated in src directory: externals, components, share, and drivers The driver_cpl directory has been moved to src/drivers/mct to prepare for multiple coupler options The share/csm_share/share directory has been renamed src/share/util. The cime script infrastructure source, utils/python/CIME, has moved to scripts/lib/CIME. @singhbalwinder : please make sure PNNL cluster merge look correct @mfdeakin-sandia : please look at cprnc/CMakeLists.txt and also confirm no run_acme changes. Fixes #1311 Fixes #1380 Fixes #1382 Fixes #1383 Fixes #1396 Fixes #1402 Fixes #1416 [BFB] * agsalin/update-to-cime5.3: (5649 commits) Fix titan problem Reset invalid pio_numiotasks Fix merge mistake Re-add eca testmod, was somehow lost in big merge Restore ACME version of shr_orb_cosz Fixed whitespace errors in Makefile killing COSP HOMME test typo fix Fixes to get tests to pass. Update ACME for new CIME directory structure More renaming. Rename ACME/cime directories in prep for CIME merge HOMME Improvement More edits to README Suggest a more portable usage of mktemp Tweak a comment Add comment on acme side Fix case for new unit_testing attribute to get_mpirun Allow run_tests.py to auto-determine the machine name fix an error make user-compset a seperate test ...
@worleyph yes, and it's on master now too. |
Not working correctly. Please reopen. I'll try to document this evening.
…Sent from my iPhone
On Apr 19, 2017, at 12:08, James Foucar ***@***.***> wrote:
@worleyph yes, and it's on master now too.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub, or mute the thread.
|
@jgfouca , with current master, with one job already submitted to the debug queue on Titan, and another job submitted to the debug queue has the following error message:
If I instead submit a non-ACME job (using qsub directly), e.g. an interactive job request, I get
Previously this error message was passed back when ./case.submit errored out. I see that something cahnged; what was ' ', is now 'None', but not sure that this is a meaningful change :-). |
@worleyph I'm seeing this too. Don't know what happened as this was working and tested on ESMCI. I'll look at it. |
@worleyph apparently there was a bad interaction between two PRs on the ESMCI side. The first fixed this issue, then a subsequent PR (also by me) messed with combining output for shell commands and re-broke it. It's now once again fixed on the ESMCI side. It should make it back to ACME fairly soon or I can duplicate the fix (risking conflicts) on the ACME side if it's urgent. |
@jgfouca , I labelled this as 'minor', but I am only one person. I can't really make the determination as to whether this can be held back until the next ESMCI sync-up. I don't know how often failures occur, and what different failure modes there are (and how important feedback is in resolving these). |
I've encountered this too, but have been able to diagnose failures from other log output, so guess not a showstopper. As long as the next update is relatively soon, can probably wait. |
@jgfouca , as I indicated in my e-mail to you and Peter Caldwell, I could use this fix now if you are willing to share. I'll just drop it into my working version. |
@worleyph , apologies, must have deleted the email from github and I didn't notice your mention. The fix is this: |
Get unit test build and run working with serial or parallel pFUnit
Update CIME to cime5.3 Update ACME to use CIME 5.3.x up to ESMCI hash 6156e0a from Thursday 4/13. scripts_regression_tests and acme_developer tests pass on penn workstation. This CIME change involves a large cosmetic change in CIME. Several directories have been renamed. The main interface to CIME, the scripts dir remains unchanged. cime/cime_config is renamed to cime/config The Fortran source code has been consolidated in src directory: externals, components, share, and drivers The driver_cpl directory has been moved to src/drivers/mct to prepare for multiple coupler options The share/csm_share/share directory has been renamed src/share/util. The cime script infrastructure source, utils/python/CIME, has moved to scripts/lib/CIME. @singhbalwinder : please make sure PNNL cluster merge look correct @mfdeakin-sandia : please look at cprnc/CMakeLists.txt and also confirm no run_acme changes. Fixes #1311 Fixes #1380 Fixes #1382 Fixes #1383 Fixes #1396 Fixes #1402 Fixes #1416 [BFB] * agsalin/update-to-cime5.3: (5649 commits) Fix titan problem Reset invalid pio_numiotasks Fix merge mistake Re-add eca testmod, was somehow lost in big merge Restore ACME version of shr_orb_cosz Fixed whitespace errors in Makefile killing COSP HOMME test typo fix Fixes to get tests to pass. Update ACME for new CIME directory structure More renaming. Rename ACME/cime directories in prep for CIME merge HOMME Improvement More edits to README Suggest a more portable usage of mktemp Tweak a comment Add comment on acme side Fix case for new unit_testing attribute to get_mpirun Allow run_tests.py to auto-determine the machine name fix an error make user-compset a seperate test ...
Update CIME to cime5.3 Update ACME to use CIME 5.3.x up to ESMCI hash 6156e0a from Thursday 4/13. scripts_regression_tests and acme_developer tests pass on penn workstation. This CIME change involves a large cosmetic change in CIME. Several directories have been renamed. The main interface to CIME, the scripts dir remains unchanged. cime/cime_config is renamed to cime/config The Fortran source code has been consolidated in src directory: externals, components, share, and drivers The driver_cpl directory has been moved to src/drivers/mct to prepare for multiple coupler options The share/csm_share/share directory has been renamed src/share/util. The cime script infrastructure source, utils/python/CIME, has moved to scripts/lib/CIME. @singhbalwinder : please make sure PNNL cluster merge look correct @mfdeakin-sandia : please look at cprnc/CMakeLists.txt and also confirm no run_acme changes. Fixes #1311 Fixes #1380 Fixes #1382 Fixes #1383 Fixes #1396 Fixes #1402 Fixes #1416 [BFB] * agsalin/update-to-cime5.3: (5649 commits) Fix titan problem Reset invalid pio_numiotasks Fix merge mistake Re-add eca testmod, was somehow lost in big merge Restore ACME version of shr_orb_cosz Fixed whitespace errors in Makefile killing COSP HOMME test typo fix Fixes to get tests to pass. Update ACME for new CIME directory structure More renaming. Rename ACME/cime directories in prep for CIME merge HOMME Improvement More edits to README Suggest a more portable usage of mktemp Tweak a comment Add comment on acme side Fix case for new unit_testing attribute to get_mpirun Allow run_tests.py to auto-determine the machine name fix an error make user-compset a seperate test ...
Update CIME to cime5.3 Update ACME to use CIME 5.3.x up to ESMCI hash 6156e0a from Thursday 4/13. scripts_regression_tests and acme_developer tests pass on penn workstation. This CIME change involves a large cosmetic change in CIME. Several directories have been renamed. The main interface to CIME, the scripts dir remains unchanged. cime/cime_config is renamed to cime/config The Fortran source code has been consolidated in src directory: externals, components, share, and drivers The driver_cpl directory has been moved to src/drivers/mct to prepare for multiple coupler options The share/csm_share/share directory has been renamed src/share/util. The cime script infrastructure source, utils/python/CIME, has moved to scripts/lib/CIME. @singhbalwinder : please make sure PNNL cluster merge look correct @mfdeakin-sandia : please look at cprnc/CMakeLists.txt and also confirm no run_acme changes. Fixes #1311 Fixes #1380 Fixes #1382 Fixes #1383 Fixes #1396 Fixes #1402 Fixes #1416 [BFB] * agsalin/update-to-cime5.3: (5649 commits) Fix titan problem Reset invalid pio_numiotasks Fix merge mistake Re-add eca testmod, was somehow lost in big merge Restore ACME version of shr_orb_cosz Fixed whitespace errors in Makefile killing COSP HOMME test typo fix Fixes to get tests to pass. Update ACME for new CIME directory structure More renaming. Rename ACME/cime directories in prep for CIME merge HOMME Improvement More edits to README Suggest a more portable usage of mktemp Tweak a comment Add comment on acme side Fix case for new unit_testing attribute to get_mpirun Allow run_tests.py to auto-determine the machine name fix an error make user-compset a seperate test ...
Get unit test build and run working with serial or parallel pFUnit
Update CIME to cime5.3 Update ACME to use CIME 5.3.x up to ESMCI hash 6156e0a from Thursday 4/13. scripts_regression_tests and acme_developer tests pass on penn workstation. This CIME change involves a large cosmetic change in CIME. Several directories have been renamed. The main interface to CIME, the scripts dir remains unchanged. cime/cime_config is renamed to cime/config The Fortran source code has been consolidated in src directory: externals, components, share, and drivers The driver_cpl directory has been moved to src/drivers/mct to prepare for multiple coupler options The share/csm_share/share directory has been renamed src/share/util. The cime script infrastructure source, utils/python/CIME, has moved to scripts/lib/CIME. @singhbalwinder : please make sure PNNL cluster merge look correct @mfdeakin-sandia : please look at cprnc/CMakeLists.txt and also confirm no run_acme changes. Fixes #1311 Fixes #1380 Fixes #1382 Fixes #1383 Fixes #1396 Fixes #1402 Fixes #1416 [BFB] * agsalin/update-to-cime5.3: (5649 commits) Fix titan problem Reset invalid pio_numiotasks Fix merge mistake Re-add eca testmod, was somehow lost in big merge Restore ACME version of shr_orb_cosz Fixed whitespace errors in Makefile killing COSP HOMME test typo fix Fixes to get tests to pass. Update ACME for new CIME directory structure More renaming. Rename ACME/cime directories in prep for CIME merge HOMME Improvement More edits to README Suggest a more portable usage of mktemp Tweak a comment Add comment on acme side Fix case for new unit_testing attribute to get_mpirun Allow run_tests.py to auto-determine the machine name fix an error make user-compset a seperate test ...
Update CIME to cime5.3 Update ACME to use CIME 5.3.x up to ESMCI hash 6156e0a from Thursday 4/13. scripts_regression_tests and acme_developer tests pass on penn workstation. This CIME change involves a large cosmetic change in CIME. Several directories have been renamed. The main interface to CIME, the scripts dir remains unchanged. cime/cime_config is renamed to cime/config The Fortran source code has been consolidated in src directory: externals, components, share, and drivers The driver_cpl directory has been moved to src/drivers/mct to prepare for multiple coupler options The share/csm_share/share directory has been renamed src/share/util. The cime script infrastructure source, utils/python/CIME, has moved to scripts/lib/CIME. @singhbalwinder : please make sure PNNL cluster merge look correct @mfdeakin-sandia : please look at cprnc/CMakeLists.txt and also confirm no run_acme changes. Fixes #1311 Fixes #1380 Fixes #1382 Fixes #1383 Fixes #1396 Fixes #1402 Fixes #1416 [BFB] * agsalin/update-to-cime5.3: (5649 commits) Fix titan problem Reset invalid pio_numiotasks Fix merge mistake Re-add eca testmod, was somehow lost in big merge Restore ACME version of shr_orb_cosz Fixed whitespace errors in Makefile killing COSP HOMME test typo fix Fixes to get tests to pass. Update ACME for new CIME directory structure More renaming. Rename ACME/cime directories in prep for CIME merge HOMME Improvement More edits to README Suggest a more portable usage of mktemp Tweak a comment Add comment on acme side Fix case for new unit_testing attribute to get_mpirun Allow run_tests.py to auto-determine the machine name fix an error make user-compset a seperate test ...
Update CIME to cime5.3 Update ACME to use CIME 5.3.x up to ESMCI hash 9802d59 from Thursday 4/13. scripts_regression_tests and acme_developer tests pass on penn workstation. This CIME change involves a large cosmetic change in CIME. Several directories have been renamed. The main interface to CIME, the scripts dir remains unchanged. cime/cime_config is renamed to cime/config The Fortran source code has been consolidated in src directory: externals, components, share, and drivers The driver_cpl directory has been moved to src/drivers/mct to prepare for multiple coupler options The share/csm_share/share directory has been renamed src/share/util. The cime script infrastructure source, utils/python/CIME, has moved to scripts/lib/CIME. @singhbalwinder : please make sure PNNL cluster merge look correct @mfdeakin-sandia : please look at cprnc/CMakeLists.txt and also confirm no run_acme changes. Fixes #1311 Fixes #1380 Fixes #1382 Fixes #1383 Fixes #1396 Fixes #1402 Fixes #1416 [BFB] * agsalin/update-to-cime5.3: (5649 commits) Fix titan problem Reset invalid pio_numiotasks Fix merge mistake Re-add eca testmod, was somehow lost in big merge Restore ACME version of shr_orb_cosz Fixed whitespace errors in Makefile killing COSP HOMME test typo fix Fixes to get tests to pass. Update ACME for new CIME directory structure More renaming. Rename ACME/cime directories in prep for CIME merge HOMME Improvement More edits to README Suggest a more portable usage of mktemp Tweak a comment Add comment on acme side Fix case for new unit_testing attribute to get_mpirun Allow run_tests.py to auto-determine the machine name fix an error make user-compset a seperate test ...
Update CIME to cime5.3 Update ACME to use CIME 5.3.x up to ESMCI hash 9802d59 from Thursday 4/13. scripts_regression_tests and acme_developer tests pass on penn workstation. This CIME change involves a large cosmetic change in CIME. Several directories have been renamed. The main interface to CIME, the scripts dir remains unchanged. cime/cime_config is renamed to cime/config The Fortran source code has been consolidated in src directory: externals, components, share, and drivers The driver_cpl directory has been moved to src/drivers/mct to prepare for multiple coupler options The share/csm_share/share directory has been renamed src/share/util. The cime script infrastructure source, utils/python/CIME, has moved to scripts/lib/CIME. @singhbalwinder : please make sure PNNL cluster merge look correct @mfdeakin-sandia : please look at cprnc/CMakeLists.txt and also confirm no run_acme changes. Fixes #1311 Fixes #1380 Fixes #1382 Fixes #1383 Fixes #1396 Fixes #1402 Fixes #1416 [BFB] * agsalin/update-to-cime5.3: (5649 commits) Fix titan problem Reset invalid pio_numiotasks Fix merge mistake Re-add eca testmod, was somehow lost in big merge Restore ACME version of shr_orb_cosz Fixed whitespace errors in Makefile killing COSP HOMME test typo fix Fixes to get tests to pass. Update ACME for new CIME directory structure More renaming. Rename ACME/cime directories in prep for CIME merge HOMME Improvement More edits to README Suggest a more portable usage of mktemp Tweak a comment Add comment on acme side Fix case for new unit_testing attribute to get_mpirun Allow run_tests.py to auto-determine the machine name fix an error make user-compset a seperate test ...
Example, if two jobs are submitted to the debug queue on Titan, the second submittsion will fail. An informative error message used to be output (CIME 5.1 and earlier), but this is now missing:
This is not the only useful error mesage that is now missing, but I don't have any other examples at the moment.
I vaguely remember already describing this problem somewhere, but can not find an existing open issue.
The text was updated successfully, but these errors were encountered: