Conversation
|
@anton-seaice, could you please add CICE5 and PIO updates? |
|
🚀 Attempted to deploy 🖥️
|
|
|
🚀 Attempted to deploy 🖥️
|
|
!redeploy |
9034c5b to
809acb8
Compare
|
🚀 Attempted to deploy 🖥️
|
|
🚀 Attempted to deploy 🖥️
|
|
!update-configs |
|
🔧 Opening Model Configuration PRs in Configurations RequestedConfigurations requested from profile
Pull Requests OpenedThe following PRs were opened: More details can be found in the workflow run: https://github.com/ACCESS-NRI/ACCESS-OM2/actions/runs/21574894826 |
|
contains
|
|
🚀 Attempted to deploy 🖥️
|
|
🚀 Attempted to deploy 🖥️
|
|
!update-configs |
|
🔧 Opening Model Configuration PRs in Configurations RequestedConfigurations requested from profile
Pull Requests OpenedThe following PRs were opened: More details can be found in the workflow run: https://github.com/ACCESS-NRI/ACCESS-OM2/actions/runs/21575395327 |
5308e1c to
3610070
Compare
|
🚀 Attempted to deploy 🖥️
|
|
🚀 Attempted to deploy 🖥️
|
a929a9a to
cb206a7
Compare
|
🚀 Attempted to deploy 🖥️
|
|
Repro tests pass for : |
|
🚀 Attempted to deploy 🖥️
|
|
CI doesn't like multiple specs I think... |
|
I thought only the version in the first line of the spec was important, we'll need to wait for Gadi |
| default: | ||
| tcl: | ||
| projections: | ||
| access-om2: '{name}/2026.02.000-{^cice5.variants.nxglob.value}x{^cice5.variants.nyglob.value}' |
There was a problem hiding this comment.
According to ACCESS-NRI/build-cd#350 we can remove the CalVer part from this string
|
Looks like we might need Tommy: ACCESS-NRI/build-cd#344 |
| - access-om2@git.2026.02.000 ^cice5 nxglob=360 nyglob=300 blckx=15 blcky=300 mxblcks=1 | ||
| - access-om2 ^cice5 nxglob=1440 nyglob=1080 blckx=30 blcky=27 mxblcks=4 | ||
| - access-om2 ^cice5 nxglob=3600 nyglob=2700 blckx=40 blcky=30 mxblcks=12 |
There was a problem hiding this comment.
| - access-om2@git.2026.02.000 ^cice5 nxglob=360 nyglob=300 blckx=15 blcky=300 mxblcks=1 | |
| - access-om2 ^cice5 nxglob=1440 nyglob=1080 blckx=30 blcky=27 mxblcks=4 | |
| - access-om2 ^cice5 nxglob=3600 nyglob=2700 blckx=40 blcky=30 mxblcks=12 | |
| - access-om2@git.2026.02.000 ^cice5 nxglob=360 nyglob=300 blckx=15 blcky=300 mxblcks=1 | |
| - access-om2@git.2026.02.000 ^cice5 nxglob=1440 nyglob=1080 blckx=30 blcky=27 mxblcks=4 | |
| - access-om2@git.2026.02.000 ^cice5 nxglob=3600 nyglob=2700 blckx=40 blcky=30 mxblcks=12 |
see #134
| default: | ||
| tcl: | ||
| projections: | ||
| access-om2: '{name}/2026.02.000-{^cice5.variants.nxglob.value}x{^cice5.variants.nyglob.value}' |
There was a problem hiding this comment.
| access-om2: '{name}/2026.02.000-{^cice5.variants.nxglob.value}x{^cice5.variants.nyglob.value}' | |
| access-om2: '{name}/{^cice5.variants.nxglob.value}x{^cice5.variants.nyglob.value}' |
|
We agreed at some point it was ok to have multiple specs where the different specs represent different processor layouts, but i can't remember if we resolved all the CI related things about it. It's not currently possible: If using: then the 'Check Spack Manifest' step fails as the specs don't have versions If using then pre-release concretisation fails as |
I don't think I agreed that we would do releases like this: I am concerned about the provenance issues and potential confusion for users. Why is it necessary to have a release with 3 different cice layouts? Pre-release I'm ok with, but this may require v8 to be merged, which contains the update to do away with the clunky version approach and use reserved definitions instead. Does it deploy to pre-release if it is a draft? |
|
We moved to spack v1.1 last week, so rather than rebase this PR - I've moved this to #135 and this can be kept for record / reference |
Next release of ACCESS-OM2:
Will bump the access-om2 version once other changes have been made.
🚀 The latest prerelease
access-om2/pr133-10at cb206a7 is here: #133 (comment) 🚀