-
Notifications
You must be signed in to change notification settings - Fork 374
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
add a pam test to gh/ci suite #6766
Comments
MMF2 tests require unreasonable amount of data that's out of sync with E3SM v3 and EAMxx v1. Thus, won't fix. |
What case were you trying to run? |
This one: SMS_Ln3_P4.ne4pg2_oQU480.F2010-MMF2.ghci-oci_gnu It tried to download the files in E3SM-Project/containers#17 missing files
... a few problems that made me abandon ship (supporting this test case long term may become a nightmare)
|
Unrelated rant: I really hope the aerosol and chemistry people rethink these files at some point... They're abjectly bad! |
@mahf708 catching up on this - the MMF2 compsets should only be using prescribed aerosol. The prognostic aerosol capability was never completed. I suspect there's something about that test modifier that's adding those aerosol files to the namelist, but I don't understand that modifier, so I'm confused how to check this. Perhaps these are the prescribed aerosol files... but that wouldn't really make sense because the MMF1 tests should be using the same files. |
If you try to build a case, you can try to find Buildconf/eam.input_data_list with the following for the test above:
I am not sure how those settings are handled though ... |
(In other words, this has nothing to do with "ghci-oci_gnu" modifier --- the quoted files above are for chrysalis_intel) |
Very likely the stuff in components/eam/cime_config/config_component.xml is inheriting random stuff from ancient cam compsets, since you don't have full-fledged files with definitions, the stuff must be coming from some ancient defaults somewhere ... |
yea - sounds like the compset parsing there is broken and the chemistry options options are being added to the CAM_CONFIG_OPTS. |
Adding this test to the gh snl tests (i.e., tests running in snl containers) will be much easier I will also note that when compiling on intel with O3, this test has a lot of non-trivial opt going on, but that's an aside |
If the issue is to protect MMF failures when scream physics is updated by a PR that does not go through next, then it makes more sense to add this test to the snl container tests.
Is this test the desired one? If so, I can add it right away, we can ninja merge, and ninja bless. Edit: I went ahead and added it. #6852 |
Will be fixed in #6852 |
Objective: avoid breaking PAM.
Background: PAM uses some EAMxx internals in specific ways that we often change.
Solution: Add one PAM/MMF2 test to gh/ci such that when specific files are touched, it is triggered to safeguard against breaking the test
The text was updated successfully, but these errors were encountered: