Open KateFriedman-NOAA opened 6 months ago
Timeline update from @aerorahul :
WAFS has a code delivery in Q1FY25 with an implementation in Q2FY25.
The GFS will be delivered and implemented at the same time removing WAFS from the GFS.
This will be v16.4.0
Just renamed the release branch to release/gfs.v16.3.21
to match the new version in ops. The ARFC for GFSv16.3.19
took the version number and the obsproc/v1.3
update will take GFSv16.3.20
. Will update release notes and internal version shortly, need to prep release branch for ARFC GFSv16.3.19
first.
Thank you @KateFriedman-NOAA for your quick action.
DA team members are meeting this afternoon to make a final decision as to what course of action to take. Submitting an ARFC to place PlanetiQ data in monitor mode is one option. Another option is to let the PlanetiQ data flow into and be assimilated by the GFS. This later option requires no EIB action.
This issue is opened as a "heads up" to EIB in case the DA team goes with the first option.
Synced the release/gfs.v16.3.21
branch with dev/gfs.v16
after the GFSv16.3.19 branch was merged into dev/gfs.v16
. Also updated the release branch for the new version v16.3.21
.
Sync merged the recently-handed-off changes for GFSv16.3.20 into the release/gfs.v16.3.21
branch and cut hand-off tag EMC-v16.3.21
.
CDF submitted to NCO.
WAFS CDF was also just submitted to NCO. EVS will also be submitting a companion CDF soon.
@YaliMao-NOAA NCO has requested that we increase the walltime of the wafs 0p25 blending job to 40mins in the new WAFSv7.0.0 package. NCO made the change locally in the GFSv16.3.20 install:
I increased the walltime of
/lfs/h1/ops/para/packages/gfs.v16.3.20/ecf/scripts/gfs/atmos/post_processing/grib2_wafs/jgfs_atmos_wafs_blending_0p25.ecf
from 30 to 40 minutes to match a change Wei made in gfs.v16.3.18. - Diane Stokes
Please make the same increase in the WAFSv7.0.0 package and recut the WAFS hand-off tag. Thanks!
@KateFriedman-NOAA Thank you for the notification. I am going to check whether it's necessary to increase the walltime in wafsv7.0.0 since blending process is parallel in wafsv7.0.0
@YaliMao-NOAA Good point, sounds good, thanks!
Merge synced the release/gfs.v16.3.20 changes into the release/gfs.v16.3.21 branch and recut the EMC-v16.3.21 tag.
This is now v16.3.22. The new ops change documented in issue #3082 will go in before this and pushes this to v16.3.22.
Description
Remove the WAFS (formerly EMC_gfs_wafs) from the GFS, it will now be a standalone system in operations.
WAFS issue: https://github.com/NOAA-EMC/WAFS/issues/23
Target version
v16.3.22
Expected workflow changes
Remove WAFS pieces:
WAFSF="NO"
-o
flag for checkout (?)