Update FATES hydro test mod to remove temporary test failure workaround #2882
Conversation
glemieux
added
test: fates
This tag includes the fix to NGEET/fates#1254 and will allow the current default fates parameter file to be used in fates hydro tests
|
@slevis-lmwg @ekluzek if we can get this (and it's FATES tag update) into master before the next b4b-dev update, I might be able to then rebase #2904 to the |
iRpointer files for restart now have the simulation timestamp in the filename Add the simulation timestamp to the rpointer files. Also update submodules with this change in CMEPS and CDEPS as well as updated cime to handle it. See the notes below for an explaination about this. Add a "clm" level directory under usermods_dirs so that the component where user-mods reside is declared and to make them function the same as test-mods.
|
Regression testing on There is one non-fates testmod that is failing build: Folder location: |
|
Regression testing on |
|
It looks like my speculation was incorrect regarding the failed runs. Re-submission did not alleviate the issues. I also attempted to generate a new baseline for All the cesm.logs are showing similar failure messages: |
|
@glemieux I'm concerned about your seeing the seg faults. But, is this just due to Izumi instability? Try the same tests with the baseline version (ctsm5.3.016) and if they fail the same way -- then let's move forward. |
|
Updated to say, try these tests for vanilla ctsm5.3.016 and if they fail the same way -- move forward. If they don't we probably need to figure it out, starting with the first tmp tag that @slevis-lmwg made. |
Roger that. Testing with |
|
Regression testing Results: |
|
From the standup this morning. The issue for @slevis-lmwg to work on is #2924. Once, we know a little more about this, we will likely let @glemieux merge this as is, since it's orthogonal and so he doesn't have to redo testing. Then @slevis-lmwg will tag a fix after that. |
There was a problem hiding this comment.
@glemieux I just looked this over. It's obviously really simple, but to confirm what's going on. The FatesHydro test used to have a special FatesHydro parameter file that it had to generate -- and now it just uses the default one? You also added a PVT test to expected fails.
Yep, that's correct. A while back, via #2700, I implemented the special FATES hydro parameter file as a temp workaround to a bug that had been exposed during testing that was on the FATES-side. The fix came in via fates tag sci.1.80.1_api.37.0.0, which we're getting via the fates |
|
@glemieux go ahead and move your baselines in place as the next tmp tag (branch_tags/tmp-241219.n01.ctsm5.3.016, and create the ChangeLog update for it, and let's make finish this off. I'm not sure who's turn it is to do the next FATES tag, so I'll volunteer to do it. |
|
Minor correction to the new tag name: |
Description of changes
This PR reverts two commits that created a workaround to hydro issue NGEET/fates#1254.
Specific notes
Contributors other than yourself, if any: @XiulinGao
CTSM Issues Fixed (include github issue #): Fixes #2878
Are answers expected to change (and if so in what way)? Yes, only for the
FatesColdHydrotestsAny User Interface Changes (namelist or namelist defaults changes)?
Does this create a need to change or add documentation? Did you do so?
Testing performed, if any: regular and fates