Fix eobs diags being left on the floor and a number of EE2 related updates#4267
Conversation
|
Testing commenced on Gaea C6 at |
Testing has concluded on GaeaC6. All tests have passed. The results are below: Marking the PR as ready for review. |
CoryMartin-NOAA
left a comment
There was a problem hiding this comment.
I assume the plan is to test through the full suite and then remove the commented out lines?
| / | ||
| / | ||
| EOF | ||
| cat fort.41 |
There was a problem hiding this comment.
is this cat necessary?
There was a problem hiding this comment.
I added this so that the stdout can show what the input namelist to this program was. It helps with debugging when things don't go as planned.
There was a problem hiding this comment.
EE2 won't like any debugging statements left in the code.
There was a problem hiding this comment.
This is not considered debugging.
|
@CatherineThomas-NOAA can you please review this PR? |
|
Except for @CoryMartin-NOAA's comment about the commented out unused variables, this PR looks good to me. Thanks @aerorahul for giving the GSI scripts some much needed TLC. I'd like to poke around the CI test directories a bit before approving. Can you share a path for a currently running/recently run CI test? Any machine is fine. |
Gaea C6 Note some tests have failed due to the libfabric error, but the COMs should be good enough for you to be able to see the changes. I can remove the unused variables, but I want to keep them as a reminder that a cleanup is needed GSI-scripting-wide, and as @CoryMartin-NOAA suggests a 1 day sprint can take care of all of them. |
|
Rewinds of the failed jobs, should succeed with the libfabric error. (Sometimes it does take multiple tries). A helpdesk ticket is open. |
Yes. That does work sometimes and with multiple tries. I did try it a few times this AM, and some attempts were successful. |
Thanks! I've poked around the COM and logs and didn't find anything unexpected.
That works for me. |
CatherineThomas-NOAA
left a comment
There was a problem hiding this comment.
Approved pending CI
|
Thank you for working on this very tricky issue @aerorahul! |
…dates (NOAA-EMC#4267) - Tar up the gsidiags into a gsidiag.tar instead of the GSI writing out to COM via symlinks. The tarred GSI diags are then untarred where they are needed. - provides a number of EE2 related updates (replacing link with copy), `COMIN/COMOUT`, etc. During the course of this , it was noticed that the EnKF was not assimilating any satellite obs. This hotfix was issued by @CatherineThomas-NOAA in a different PR already. Several variables in the analysis-related scripts are unnecessary and seem to come from earlier copy/paste versions. They can lead to confusion. This PR tries to eliminate some, though more work is recommended before code hand-off to expedite code acceptance into ops. There are also several deprecated "options" that are either no longer functional or not used.
Description
This PR:
COMIN/COMOUT, etc.During the course of this , it was noticed that the EnKF was not assimilating any satellite obs. This hotfix was issued by @CatherineThomas-NOAA in a different PR already.
Several variables in the analysis-related scripts are unnecessary and seem to come from earlier copy/paste versions. They can lead to confusion. This PR tries to eliminate some, though more work is recommended before code hand-off to expedite code acceptance into ops. There are also several deprecated "options" that are either no longer functional or not used.
Type of change
Change characteristics
How has this been tested?
Tested with the C96C48_hybatmDA test on Gaea.
Full testing is in progress on GaeaC6 and can be viewed at:
/gpfs/f6/drsa-precip4/world-shared/Rahul.Mahajan/RUNTESTS/gw-2092Checklist