MPI reproducibility fix in w3scat #518
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
sea points that recently were covered with ice (IC0)
|
The tests ran successfully with pre known nonb4b cases. The changes in ww3_ufs? and mww3_test_01 cases and unstructured cases were expected. |
aliabdolali
approved these changes
Nov 5, 2021
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Pull Request Summary
The w3scat subroutine is updated so that MAPSTA.ne.0 VA values are updated.
Description
To have two runs of ww3_multi or ww3_shel with different number of MPI tasks should produce the same result. However, it was found that when using for example IC0 and a sea point (MAPSTA=1) is deactivated because the ice concentration is becomes greater than FICEN and then then the point is deactivated (MAPSTA=-1). This point should have VA=0 then, but when using MPI it was not correctly 0. However, if we update the call in w3scat to update not just for MAPSTA >0 but MAPSTA .ne.0 then, we get the same results.
Issue(s) addressed
Check list
Is your feature branch up to date with the authoritative repository (NOAA/develop)? yes
Make sure you have checked the checklist for a developer submitting to develop and updating version number
Please list appropriate labels code managers should add for this PR:
bug
Reviewers: @aliabdolali @MatthewMasarik-NOAA
Commit Message
Testing
Difference with develop: These have changes in the expected not b4b tests plus there are differences in tests where the change now produces b4b results when running with different number of tasks for MPI.
matrixCompSummary.txt
matrixCompFull.txt
matrixDiff.txt
These are the differences when running this branch but with two different MPI counts. Note that (with the exception of log files) the ufs tests and mww3_test_01 now are reproducible, so the change above is good.
matrixDiff.txt
matrixCompFull.txt
matrixCompSummary.txt
The changes for unstructured grid are thought to be solving an issue, but instead of ice the impact is with changing of water level. The changes in answers @aliabdolali has found to be very small. However, it seems that there are still some outstanding issues with MPI reproducibility and unstructured grids that this fix alone did not solve.