Skip to content
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.

Sign up for GitHub

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

Jump to bottom

Reduce number of CCPP suites after ufs-weather-model regression test update #282

Closed
climbfuji opened this issue Apr 15, 2021 · 9 comments
Closed

Reduce number of CCPP suites after ufs-weather-model regression test update #282

climbfuji opened this issue Apr 15, 2021 · 9 comments
Assignees
@climbfuji
Labels
enhancement New feature or request

Comments

@climbfuji
Copy link
Collaborator

Description

We currently (2021/04/15) have 68 suites in FV3's ccpp/suites directory. It is tedious to make changes to CCPP schemes that then require changing each suite definition file.

Solution

After the update of the ufs-weather-model regression tests has been completed, I suggest:

  1. Identify which suites are used by the regression tests, keep those.
  2. For the remaining suites, identify owners and whether these suites are still actively used in development and testing by those individuals or collaborators. If yes, keep, if no, delete.
  3. Rinse and repeat every year.

Alternatives

Let the number of suite definition file grow and grow.

Testing:

n/a

Dependent PRs:

n/a
@climbfuji climbfuji added the enhancement New feature or request label Apr 15, 2021
@climbfuji
Copy link
Collaborator Author

As of 2021/08/02, we have 80 suites in directory ccpp/suites. The ufs-weather-model regression tests use 41 of these. I believe that some of the following suites that are not used can be deleted. I would like to ask the physics developers and fv3atm code managers to help me identify which suites are no longer needed. @yangfanglin @SMoorthi-emc @junwang-noaa @DusanJovic-NOAA @ligiabernardet

I created a spreadsheet to keep track of what to keep and delete, this is the link to VIEW the spreadsheet: https://docs.google.com/spreadsheets/d/1oDF_L1pkt-J1engAnD431FhE-q09XtZ78LvrT4wZ1Y8/edit?usp=sharing

If you want edit/comment access, please let me know. Thanks!

Suite FV3_CPT_v0 NOT used
Suite FV3_GFS_2017 NOT used
Suite FV3_GFS_2017_coupled NOT used
Suite FV3_GFS_2017_couplednsst NOT used
Suite FV3_GFS_2017_csawmg NOT used
Suite FV3_GFS_2017_gfdlmp NOT used
Suite FV3_GFS_2017_gfdlmp_noahmp NOT used
Suite FV3_GFS_2017_gfdlmp_regional NOT used
Suite FV3_GFS_2017_h2ophys NOT used
Suite FV3_GFS_2017_ozphys_2015 NOT used
Suite FV3_GFS_2017_sas NOT used
Suite FV3_GFS_2017_satmedmf NOT used
Suite FV3_GFS_2017_satmedmf_coupled NOT used
Suite FV3_GFS_2017_satmedmfq NOT used
Suite FV3_GFS_cpld_rasmgshoc NOT used
Suite FV3_GFS_cpld_rasmgshocnsst NOT used
Suite FV3_GFS_cpld_rasmgshocnsst_flake NOT used
Suite FV3_GFS_cpldnst_rasmgshoc NOT used
Suite FV3_GFS_rasmgshoc NOT used
Suite FV3_GFS_v15 NOT used
Suite FV3_GFS_v15_ras NOT used
Suite FV3_GFS_v15_rasmgshoc NOT used
Suite FV3_GFS_v15_thompson NOT used
Suite FV3_GFS_v15_thompson_mynn_lam3km NOT used
Suite FV3_GFS_v15p2 NOT used
Suite FV3_GFS_v15p2_RRTMGP NOT used
Suite FV3_GFS_v15p2_coupled NOT used
Suite FV3_GFS_v15p2_couplednsst NOT used
Suite FV3_GFS_v15p2_no_nsst NOT used
Suite FV3_GFS_v15plus NOT used
Suite FV3_GFS_v15plusras NOT used
Suite FV3_GFS_v16_coupled_nsstNoahmp NOT used
Suite FV3_GFS_v16_no_nsst NOT used
Suite FV3_GFSv17alpha_cpldnsstras NOT used
Suite FV3_GFSv17alpha_cpldnsstras_flake NOT used
Suite FV3_GFSv17alpha_cpldnsstrasnoshal NOT used
Suite FV3_GFSv17alpha_cpldnsstsas NOT used
Suite FV3_GFSv17alpha_ras NOT used
Suite FV3_GFSv17alpha_ras_flake NOT used
Suite FV3_GFSv17alpha_sas NOT used
Suite FV3_GSD_v0_unified_ugwp_suite_noah NOT used

@climbfuji climbfuji self-assigned this Aug 2, 2021
@yangfanglin
Copy link
Collaborator

Dom, when you say the following suites that are not used can be deleted, do you mean they are not used by any of the regression tests ? Can you give all access the spreadsheet to make comments ?

@SMoorthi-emc
Copy link
Contributor

SMoorthi-emc commented Aug 2, 2021 via email

@climbfuji
Copy link
Collaborator Author

Dom, when you say the following suites that are not used can be deleted, do you mean they are not used by any of the regression tests ? Can you give all access the spreadsheet to make comments ?

Yes, they are not used by any of the regression tests. I know that several suites should be kept for testing by our developers, but there are several that won't be used anymore (I would assume some of the old GFS v14 suites). This is why I created the spreadsheet so that we can gather information. It's open for commenting for all NOAA folks.

@climbfuji
Copy link
Collaborator Author

What is the harm in keeping SDFs? Considering that CCPP requires different SDF to turn on different physics options we should encourage having more SDF not less. I am very disappointed with thought of having to remove SDF. This is not helping community modeling. Sorry, I don’t endorse this. Moorthi
…

Please see my reply to Fanglin. I think there are several suites that are really no longer in use, for example FV3_CPT_v0 or some of the older GFS v14 variants. I am not saying we should remove suites that people are experimenting with (although a point can be made that these could be kept in user forks, but I know this is cumbersome and therefore prefer not to go this way).

But please keep in mind that the more suites you have, the more housekeeping is needed. And it would be nice if we could compile all suites (by not specifying the argument CCPP_SUITES=...). This still works with GNU, but no longer with Intel (internal compiler error, tries to optimize too much). It used to work with Intel when we had about 40 suites or so.

@climbfuji climbfuji mentioned this issue Aug 11, 2021
Merged
@climbfuji
Copy link
Collaborator Author

@DeniseWorthen Last I remember you were trying to work on that using the spreadsheet https://docs.google.com/spreadsheets/d/1oDF_L1pkt-J1engAnD431FhE-q09XtZ78LvrT4wZ1Y8/edit?usp=sharing. Can you remind me about the current status, please?

@DeniseWorthen
Copy link
Collaborator

@climbfuji I am planning on removing FV3_GFS_2017_coupled and suite_FV3_GFS_2017_satmedmf_coupled.xml. I can't do this until the PR #765 has been committed.

@DeniseWorthen DeniseWorthen mentioned this issue Sep 22, 2021
Closed
@arunchawla-NOAA
Copy link

@climbfuji

can we close this issue ? the PR associated with this issue is already closed

@climbfuji
Copy link
Collaborator Author

Yes ... I am declaring defeat on my overly ambitious goal of really reducing the number of suites ;-)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@climbfuji climbfuji

Labels
enhancement New feature or request
Projects
None yet
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

Remove unused coupled suites DeniseWorthen/fv3atm
5 participants
@climbfuji @DeniseWorthen @yangfanglin @SMoorthi-emc @arunchawla-NOAA