Closed guillaumevernieres closed 1 year ago
@guillaumevernieres you are right on this. Originally we got the fixed files from Travis and thought it is 4x4. A simple renaming will cause the job failure for users working branch. A viable way is to copy them to 5x5 naming style while keep 4x4 on disk for a period of time. In the meantime we need to have a simple PR to correct the fixed file directory in rt.sh. After the PR is done, we can ask users to sync their working branch and delete the original 4x4 fixed file directory.
Thanks for your input @jiandewang . @KateFriedman-NOAA , could you make a copy instead of renaming as @jiandewang suggested?
@KateFriedman-NOAA can you do the coping on all platforms as I don't have permission on most of them ? @WalterKolczynski-NOAA Can you also assign this issue to me? I am going to have a PR on script changing for this.
@jiandewang I think I can just fix it quickly.
@jiandewang I think I can just fix it quickly.
sounds good
I have fixed this in the fix set on Hera and am currently rsyncing the change to the other platform fix sets:
[glopara@hfe05 20220805]$ pwd
/scratch1/NCEPDEV/global/glopara/fix/orog/20220805
[glopara@hfe05 20220805]$ ll C48.mx500_frac
total 190568
-rw-r--r-- 1 glopara global 539002 May 4 01:39 C48_grid.tile1.nc
-rw-r--r-- 1 glopara global 539002 May 4 01:39 C48_grid.tile2.nc
-rw-r--r-- 1 glopara global 539002 May 4 01:39 C48_grid.tile3.nc
-rw-r--r-- 1 glopara global 539002 May 4 01:39 C48_grid.tile4.nc
-rw-r--r-- 1 glopara global 539002 May 4 01:39 C48_grid.tile5.nc
-rw-r--r-- 1 glopara global 539002 May 4 01:39 C48_grid.tile6.nc
-rw-r--r-- 1 glopara global 22010 May 4 01:39 C48_mosaic.nc
drwxr-sr-x 2 glopara global 4096 May 4 01:55 fix_sfc
-rw-r--r-- 1 glopara global 3459336 May 4 01:39 fv3_SCRIP_C48_GRIDSPEC_lon192_lat94.gaussian.bilinear.nc
-rw-r--r-- 1 glopara global 2593056 May 4 01:39 fv3_SCRIP_C48_GRIDSPEC_lon192_lat94.gaussian.neareststod.nc
-rw-r--r-- 1 glopara global 226627 May 4 01:50 oro_C48.mx500.tile1.nc
-rw-r--r-- 1 glopara global 226627 May 4 01:50 oro_C48.mx500.tile2.nc
-rw-r--r-- 1 glopara global 226627 May 4 01:50 oro_C48.mx500.tile3.nc
-rw-r--r-- 1 glopara global 226627 May 4 01:50 oro_C48.mx500.tile4.nc
-rw-r--r-- 1 glopara global 226627 May 4 01:50 oro_C48.mx500.tile5.nc
-rw-r--r-- 1 glopara global 226627 May 4 01:50 oro_C48.mx500.tile6.nc
-rw-r--r-- 1 glopara global 131815449 May 4 01:39 remap_weights_C48_0p125deg.nc
-rw-r--r-- 1 glopara global 37056281 May 4 01:39 remap_weights_C48_0p25deg.nc
-rw-r--r-- 1 glopara global 11457561 May 4 01:39 remap_weights_C48_0p5deg.nc
-rw-r--r-- 1 glopara global 4096473 May 4 01:39 remap_weights_C48_1deg.nc
@WalterKolczynski-NOAA is working on the associated workflow update.
Thanks @KateFriedman-NOAA .
@guillaumevernieres The fix sets on Hera, Orion, Cactus, and Dogwood are updated. Waiting for Jet to come back from maintenance to update the set there.
Jet fix set is now also updated.
The bug
The orography fixed files for the C48/5 degree MOM6-CICE are not named properly on
Hera
(probably true for other HPC):400
should be replaced by500
. As far as I know there isn't an existing 4 degree MOM6/CICE. Correct me if I am wrong @jiandewang , @DeniseWorthen or @JessicaMeixner-NOAA.The solution
/scratch1/NCEPDEV/global/glopara/fix/orog/20220805/C48.mx400_frac
should be/scratch1/NCEPDEV/global/glopara/fix/orog/20220805/C48.mx500_frac
and
oro_C48.mx400.tile*.nc
should beoro_C48.mx500.tile*.nc