Updating modulefiles to work with latest spack-stack #501
Conversation
|
Do the modulefile changes correspond to the latest in https://github.com/ufs-community/ufs-weather-model/tree/develop/modulefiles? |
|
@grantfirl Yes, these changes should correspond to the latest weather model modules (spack-stack 1.6.0) |
There was a problem hiding this comment.
@mkavulich These changes look good. Just a small comment on a documentation change that I think should be reverted.
scm/doc/TechGuide/chap_quick.rst
Outdated
| @@ -491,10 +501,11 @@ To see the full list of available options, use the ``--help`` flag: | |||
|
|
|||
| The run script’s full set of options are described below, where optional abbreviations are included in brackets. | |||
| If using the main branch, you should run the above command to ensure you have the most up-to-date list of options. | |||
| There are no required arguments, but at least one of ``--case`` or ``--file`` must be specified. | |||
There was a problem hiding this comment.
This sentence is confusing. I would revert this change.
Can you add a sentence below that if no --suite is selected, the default SDF will be used? Which is GFS_V16
There was a problem hiding this comment.
I don't quite understand why this is confusing, is it the use of the term "required argument"? The script right now fails in a non-intuitive way if users don't specify one of these arguments, so I wanted to add a note about that specific requirement.
Maybe a better way forward is to just fix the script so it gives a clear error message when it doesn't get an argument for --file or --case, does that sound better?
There was a problem hiding this comment.
It's confusing, at least to me, that there are (a) no required arguments, but (b) --case or --file must be specified as an argument.
But yeah, I agree that the better approach is to modify run_scm.py to report error if (not args.case) and (not args.file), and reflect that here in the docs (plus the bit about the default suite).
There was a problem hiding this comment.
Thanks for the feedback; I pushed an update, let me know what you think.
mkavulich
force-pushed
the
feature/modulefile_updates
branch
from
a2b6b09
to
adca29b
Compare
|
@mkavulich Can you pull in the latest main before we merge this, please? |
…orking and they arent necessary)
mkavulich
force-pushed
the
feature/modulefile_updates
branch
from
adca29b
to
0e51f53
Compare
The modulefiles have not been updated in a while and they do not work on some platforms. This PR updates the modulefiles for the latest spack-stack installations (1.6.0) on Derecho, Hera, Jet, and Orion; all the existing. It removes the Jet and Orion GNU modulefiles (I don't think they ever actually worked, these are not officially supported installations). Also, just a few fixes and clarifications for the users guide about spack-stack.
Successfully compiled on Derecho (intel and GNU), Hera (intel and GNU), Jet (Intel) and Orion (Intel). Many regression tests are failing at runtime (35/50), but this is a known problem that will be fixed in other PRs.
Finally, a fix to the supported suites list to remove SCM_RAP (which is no longer supported).