Support GFSV16.2.0 on Hera/Orion

[WenMeng-NOAA]

The changes for support GFS V16.2.0 on Hera/Orion:

• Use NCEPLIBs built from hpc/1.2.0.
• Use modulefiles in lua format

[KateFriedman-NOAA]

Updates to support Hera/Orion for GFSv16.2.0 look good. I tested building the post_gfsv16_hpc branch on both Hera and Orion, please see the following logs to confirm they built as expected with the g-w versions files in play:
Hera: /scratch1/NCEPDEV/global/Kate.Friedman/git/dev_v16/sorc/logs/build_ncep_post.log_post_gfsv16_hpc
Orion: /work/noaa/global/kfriedma/git/dev_v16/sorc/logs/build_ncep_post.log_post_gfsv16_hpc

[WenMeng-NOAA]

@KateFriedman-NOAA Both build logs look good to me. When you may set "machine=ORION" for orion and "machine=HERA" for hera in g-w runtime configuration. When this PR merged, I will create a new tag upp_v8.1.1 for supporting GFS V16.2.0 on hera and orion. Please let me know if you have comments on tag name.

[KateFriedman-NOAA]

Both build logs look good to me

Excellent, thanks for confirming!

When you may set "machine=ORION" for orion and "machine=HERA" for hera in g-w runtime configuration.

Not sure I understand what you're saying here, can you elaborate? Thanks!

new tag upp_v8.1.1 for supporting GFS V16.2.0 on hera and orion

Tag name works for me! I'll be testing this tag (and the other updated components for GFSv16.2.0 R&D support) on WCOSS2 to confirm the R&D updates do not break/impact WCOSS2. Is there any concern regarding WCOSS2? Thanks!

[WenMeng-NOAA]

@KateFriedman-NOAA

When you may set "machine=ORION" for orion and "machine=HERA" for hera in g-w runtime configuration.

Not sure I understand what you're saying here, can you elaborate? Thanks!

It's just a reminder. The undefining or setting incorrectly variable "machine" from upstream configuration might cause MPMD process unlaunched. That would take a longer time to generate PGRB datasets.

[KateFriedman-NOAA]

When you may set "machine=ORION" for orion and "machine=HERA" for hera in g-w runtime configuration.
Not sure I understand what you're saying here, can you elaborate? Thanks!

It's just a reminder. The undefining or setting incorrectly variable "machine" from upstream configuration might cause MPMD process unlaunched. That would take a longer time to generate PGRB datasets.

Oh ok, I see. So the global-workflow setup scripts define machine at runtime (in config.base and the rocoto xml) as one of the following: WCOSS2, HERA, ORION, or WCOSS_DELL_P3. We'll have JET and possibly others down the road too. Those settings line up with what you state is used by the UPP so I think we're good. At build-time the global-workflow machine-setup.sh script sets a target variable that is the lowercase equivalent of machine but the characters are the same. Is there any issue with the lowercase target variable at buildtime for the UPP?

[WenMeng-NOAA]

@KateFriedman-NOAA Thanks for confirming from g-w end. The UPP build script sorc/ uses lowercase for "machine" setting in build time. There should no issue for build time in UPP standalone and global-work perspective so far.

[KateFriedman-NOAA]

The UPP build script sorc/ uses lowercase for "machine" setting in build time. There should no issue for build time in UPP standalone and global-work perspective so far.

Good, thanks for confirming the machine variable usage on the UPP side for build time. Sounds like we're good for both run and build.