Remove need for a ice_prod dependency check script #2809
Conversation
The linters.yaml has been updated so that shellnorms can run whenever a push occurs to EricSinsky-NOAA/develop.
The changes to linters.yaml have been reverted.
The check_ice_netcdf dependency has been removed for the ice_prod task. The rocoto age datadep dependency has been restored for the ice_prod task.
The check_ice_netcdf script has been deleted.
A function has been added to postdet to calculate a CICE average from 6-hour averaged CICE forecast data.
The cdo utility is used to perform the time averaging in the cice_avg function.
The FHOUT_ICE_6h has been removed and has been replaced with the already defined variable FHOUT_ICE.
![github-advanced-security[bot]](https://avatars.githubusercontent.com/in/57789?s=60&v=4){kind=small}
| echo "SUB ${FUNCNAME[0]}: No further averaging needed for CICE" | ||
| else | ||
| echo "SUB ${FUNCNAME[0]}: Averaging output data for CICE" | ||
| CDO=/apps/prod/hpc-stack/intel-19.1.3.304/cdo/1.9.8/bin/cdo |
There was a problem hiding this comment.
If we choose to use the cdo utility for this, it may need to be added somewhere in jobs/JGLOBAL_FORECAST. This line is just a temporary workaround for the cdo utility being undefined in the forecast task environment.
There was a problem hiding this comment.
The cdo binary directory is already added to the path by the cdo module. You should be able to just call it with cdo without this.
There was a problem hiding this comment.
NB: We have it in the base modulefile used for most jobs, but the model modulefile isn't loading cdo, so it would need to be added to load_ufswm_modules.sh.
There was a problem hiding this comment.
Thanks, @WalterKolczynski-NOAA. Will module load cdo in load_ufswm_modules.sh.
emcbot
added
CI-Orion-Running
|
Experiment C48_ATM FAILED on Orion in Build# 2 with error logs: Follow link here to view the contents of the above file(s): (link) |
emcbot
added
CI-Orion-Failed
|
Experiment C48_ATM FAILED on Orion in Build# 2 in |
|
Experiment C48_S2SWA_gefs FAILED on Orion in Build# 2 in |
|
Experiment C48_S2SW FAILED on Orion in Build# 2 with error logs: Follow link here to view the contents of the above file(s): (link) |
|
Experiment C48_S2SW FAILED on Orion in Build# 2 in |
|
Experiment C96C48_hybatmDA FAILED on Orion in Build# 2 with error logs: Follow link here to view the contents of the above file(s): (link) |
|
Experiment C96C48_hybatmDA FAILED on Orion in Build# 2 in |
|
Ignore the Orion failure. We already know what that is. |
WalterKolczynski-NOAA
removed
the
CI-Orion-Failed
|
Experiment C96_atm3DVar FAILED on Orion in Build# 2 with error logs: Follow link here to view the contents of the above file(s): (link) |
emcbot
added
the
CI-Orion-Failed
|
Experiment C96_atm3DVar FAILED on Orion in Build# 2 in |
emcbot
added
CI-Orion-Failed
|
CI Failed on Orion in Build# 2 |
emcbot
added
CI-Wcoss2-Passed
|
All CI Test Cases Passed on Wcoss2: |
WalterKolczynski-NOAA
removed
the
CI-Orion-Failed
emcbot
added
CI-Hera-Passed
|
CI Passed on Hera in Build# 3 |
5 participants
Description
The main purpose of this PR is to remove the need for an ice_prod dependency check script
ush/check_ice_netcdf.sh. The original purpose of the ice_prod dependency check script is to check for special case dependencies where( cyc + FHMIN ) % FHOUT_ICE )) =! 0(more details on this issue can be found in issue #2674 ). A bugfix for these special cases is expected to come from a PR in the ufs-weather-model.Resolves #2721
Refs #2721, #2674
Type of change
Change characteristics
How has this been tested?
FHOUT_OCNandFHOUT_ICEwas set to24and the GEFS CI case was successfully tested (202103232 12Z).FHOUT_OCNandFHOUT_ICEset to6.Checklist