Absorb GLDAS scripts #1014
Comments
KateFriedman-NOAA
added
feature
|
I don't think there will be any issue as long as the right path is given in the scripts. |
Great! Yep, I'll be making the necessary workflow-side updates related to paths and such. :) |
- Remove GLDAS JJOB script from ignore file. - Remove GLDAS exscript from ignore file. - Remove GLDAS ush scripts from ignore file. Refs NOAA-EMC#1014
- Remove symlink creation for GLDAS scripts in the workflow link script. - Removal includes JJOB, ex-script, and ush script symlinks. Refs NOAA-EMC#1014
- Initial copy of GLDAS JJOB, exscript, and ush scripts from GLDAS repo into global-workflow. Refs NOAA-EMC#1014
- Convert from ksh to bash. - Convert backticks to substitutions. - Add preamble. - Remove announcements and general cleanup. Refs NOAA-EMC#1014
- Remove references to WCOSS_DELL_P3 and WCOSS_C in GLDAS scripts. - Update $HOMEgfs variable in GLDAS preable source lines to address warnings from linter regarding: 1) HOMEgfs is referenced but not assigned. 2) Prefer putting braces around variable references even when not strictly required. Refs NOAA-EMC#1014
- Address SC2016 (double instead of single quote for ERRSCRIPT and SC2250 (curly braces). Refs NOAA-EMC#1014
|
@HelinWei-NOAA It does not appear that |
- Mostly address "referenced but not assigned" warnings Refs NOAA-EMC#1014
|
@KateFriedman-NOAA Both scripts are invoked by exgdas_atmos_gldas.sh. gldas_archive.sh is turn on only for offline gldas run. But I am surprise that gldas_post.sh isn't invoked at runtime. Can you point me to your log file running gldas? Thanks. |
Good to know, thanks!
I checked both the ops 00z gldas job logs and my own testing logs: Ops logs from 00z last night (/lfs/h1/ops/prod/output/20220914/gdas_atmos_gldas_00.o19964486): My test logs on Cactus: |
|
@KateFriedman-NOAA If you just search gldas_post, you can find one in the log. line 114656 of gdas_atmos_gldas_00.o19964486 This is part of gldas_post.sh. So it is invoked at runtime. |
|
The |
|
@KateFriedman-NOAA You are right. I can confirm that gldas_post.sh is not invoked at runtime. Thanks. |
|
Thanks @HelinWei-NOAA ! Since Related question...how does one run GLDAS in offline mode and invoke |
|
There is no reason to keep gldas_post.sh. We need to set OFFLINE_GLDAS = "YES" to invoke gldas_arch.sh in the offline driver submit_test_gdas_gldas_dell.sh. We haven't updated this driver for a long time. |
Ok, thanks @HelinWei-NOAA for confirming! I will delete it from my global-workflow branch and not include it in the GLDAS script absorption.
Gotcha, thanks for explaining! Since Let me know if you have any objections to the above. Thanks! |
After discussion with GLDAS code manager and confirming what GLDAS scripts are actually used at runtime, it was decided to not bring the ush/gldas_archive.sh and ush/gldas_post.sh scripts from GLDAS into global-workflow. Deleting them from this branch. Refs NOAA-EMC#1014
- Add USE_CFP and CFP_MP settings into gldas block in WCOSS2.env. - When on WCOSS2 the gldas job should use CFP (USE_CFP=YES) but not use the line counter value in the generated cfile (CFP_MP=NO). Related to work happening in the GLDAS script absorption issue NOAA-EMC#1014. Refs NOAA-EMC#419
- Remove checks for machine in ush/gldas_get_data.sh and replace with existing CFP variables in system (USE_CFP and CFP_MP). - Update Hera and Orion env files to set USE_CFP and CFP_MP as needed. - Similar env variable settings will be added into WCOSS2.env in that branch. Refs NOAA-EMC#1014
- Updates to set GLDAS fix files based on FRAC_GRID setting. - Update scripts/exgdas_atmos_gldas.sh to set FIXgldas subfolder based on value of FRAC_GRID: use "frac_grid" subfolder when .true. or use "nonfrac_grid" subfolder when .false. - Move FRAC_GRID variable from config.fcst to config.base since it will now be used in more than just the forecast job. - Also correct FIXgldas default in scripts/exgdas_atmos_gldas.sh to add missing "/fix_gldas" path. Refs NOAA-EMC#1014
…a.sh - Update GLDAS scripts and config.fcst to address linter/shellcheck warnings. - Update gldas_get_data.sh and gldas_process_data.sh to pass $iter variable as argument in gldas_process_data.sh and give to preamble. Refs NOAA-EMC#1014
- Add missing indentation into ush/gldas_forcing.sh and ush/gldas_get_data.sh. Refs NOAA-EMC#1014
- Provide a value of "1" for new exit statements in GLDAS JJOB script. Refs NOAA-EMC#1014
Absorb GLDAS scripts into global-workflow and fix GLDAS job by updating scripts to use new GLDAS fix file set. * Remove GLDAS scripts from .gitignore * Remove GLDAS script symlinks from link_workflow.sh * Add GLDAS scripts to global-workflow * Updates to GLDAS scripts, includes converting GLDAS script to replace machine checks with CFP variables * Address linter warnings and remove obsolete platforms Refs #622 #1014
Description
As part of the overarching goal of moving global workflow scripts out of component repositories and under global workflow, all of the scripts currently linked in GLDAS should be moved to global-workflow. FYI @HelinWei-NOAA , please let us know of any concerns, questions, or objections. Once we absorb the following scripts you can remove them in your GLDAS repo.
Scripts to be absorbed:
gldas_archive.sh(not bringing into global-workflow, see discussion below Absorb GLDAS scripts #1014 (comment))gldas_post.sh(not bringing into global-workflow, see discussion below Absorb GLDAS scripts #1014 (comment))Requirements
All GLDAS jobs, scripts, and ush scripts that are currently linked by global workflow added to the global workflow repo instead and removed from the ignore list.
Acceptance Criteria (Definition of Done)
No jobs, scripts, or ush scripts linked from GLDAS.
Dependencies
None
The text was updated successfully, but these errors were encountered: