Merge branch 'main' into glwagner-patch-1

Author: Sbozzolo

.buildkite/Manifest-v1.11.toml

@@ -1,5 +1,5 @@
 agents:
-  modules: "climacommon/2024_12_16"
+  modules: "climacommon/2025_03_18"
   queue: "new-central"
 steps:
   - label: "init :computer:"

.buildkite/cpu_scaling_pipeline/pipeline.yml

@@ -1,7 +1,7 @@
 agents:
   queue: clima
   slurm_mem: 8G
-  modules: climacommon/2024_12_16 nsight-systems/2024.6.1
+  modules: climacommon/2025_03_18 nsight-systems/2024.6.1
 
 env:
   JULIA_MPI_HAS_CUDA: "true"

.buildkite/gpu_pipeline/pipeline.yml

@@ -1,7 +1,7 @@
 agents:
   queue: clima
   slurm_mem: 8G
-  modules: climacommon/2024_12_16 nsight-systems/2024.6.1
+  modules: climacommon/2025_03_18 nsight-systems/2024.6.1
 
 env:
   JULIA_MPI_HAS_CUDA: "true"
@@ -140,7 +140,7 @@ steps:
           CLIMACOMMS_DEVICE: "CUDA"
           JOB_NAME: "longrun_aquaplanet_allsky_diagedmf_0M"
 
-      - label: ":computer: aquaplanet equilmoist allsky radiation + prognostic edmf diffusion only + 0M microphysics"
+      - label: ":computer: aquaplanet equilmoist allsky radiation + prognostic edmf + 0M microphysics"
         command:
           - srun julia --project=.buildkite .buildkite/ci_driver.jl --config_file $CONFIG_PATH/$$JOB_NAME.yml --job_id $$JOB_NAME
         artifact_paths: "$$JOB_NAME/output_active/*"
@@ -149,7 +149,7 @@ steps:
           slurm_time: 24:00:00
         env:
           CLIMACOMMS_DEVICE: "CUDA"
-          JOB_NAME: "longrun_aquaplanet_allsky_progedmf_diffonly_0M"
+          JOB_NAME: "longrun_aquaplanet_allsky_progedmf_0M"
 
       - label: ":computer: aquaplanet equilmoist allsky radiation + 0M microphysics + earth topography"
         command:

.buildkite/longruns_gpu/pipeline.yml

@@ -1,7 +1,7 @@
 agents:
   queue: new-central
   slurm_mem: 8G
-  modules: climacommon/2024_12_16
+  modules: climacommon/2025_03_18
 
 env:
   JULIA_LOAD_PATH: "${JULIA_LOAD_PATH}:${BUILDKITE_BUILD_CHECKOUT_PATH}/.buildkite"
@@ -23,16 +23,14 @@ steps:
     command:
       - "echo $$JULIA_DEPOT_PATH"
 
+      - echo "--- Check that ClimaAtmos is developed in Manifest"
+      - julia --check-bounds=yes -e 'using TOML; @assert TOML.parsefile(".buildkite/Manifest-v1.11.toml")["deps"]["ClimaAtmos"][1]["path"] == ".."'
+
       - echo "--- Instantiate project"
-      - "julia --project -e 'using Pkg; Pkg.instantiate(;verbose=true)'"
-      - "julia --project -e 'using Pkg; Pkg.precompile()'"
-      - "julia --project -e 'using Pkg; Pkg.status()'"
+      - "julia --project -e 'using Pkg; Pkg.instantiate(;verbose=true); Pkg.precompile(); Pkg.status()'"
 
       - echo "--- Instantiate .buildkite"
-      - "julia --project=.buildkite -e 'using Pkg; Pkg.instantiate(;verbose=true)'"
-      - "julia --project=.buildkite -e 'using Pkg; Pkg.precompile()'"
-      - "julia --project=.buildkite -e 'using CUDA; CUDA.precompile_runtime()'"
-      - "julia --project=.buildkite -e 'using Pkg; Pkg.status()'"
+      - "julia --project=.buildkite -e 'using Pkg; Pkg.instantiate(;verbose=true); Pkg.precompile(;strict=true); using CUDA; CUDA.precompile_runtime(); Pkg.status()'"
 
     agents:
       slurm_cpus_per_task: 8
@@ -136,6 +134,8 @@ steps:
           --job_id single_column_hydrostatic_balance_ft64
           --out_dir single_column_hydrostatic_balance_ft64/output_active
         artifact_paths: "single_column_hydrostatic_balance_ft64/output_active/*"
+        agents:
+          slurm_constraint: icelake|cascadelake|skylake|epyc
 
   - group: "Box Examples"
     steps:
@@ -365,6 +365,7 @@ steps:
         artifact_paths: "aquaplanet_nonequil_allsky_gw_res/output_active/*"
         agents:
           slurm_mem: 20GB
+          slurm_constraint: icelake|cascadelake|skylake|epyc
 
       - label: ":computer: aquaplanet equil allsky monin_obukhov varying insol gravity wave (raw_topo) high top zonally asymmetric"
         command: >
@@ -397,7 +398,7 @@ steps:
         artifact_paths: "aquaplanet_equil_clearsky_tvinsol_0M_slabocean_ft64/output_active/*"
         agents:
           slurm_mem: 20GB
-      
+
       - label: ":computer: aquaplanet for coupler"
         command: >
           julia --color=yes --project=.buildkite .buildkite/ci_driver.jl
@@ -587,6 +588,15 @@ steps:
         agents:
           slurm_mem: 20GB
 
+      - label: ":man_in_business_suit_levitating: AMIP Target EDOnly nonequil"
+        command: >
+          julia --color=yes --project=.buildkite .buildkite/ci_driver.jl
+          --config_file $CONFIG_PATH/amip_target_edonly_nonequil.yml
+          --job_id amip_target_edonly_nonequil
+        artifact_paths: "amip_target_edonly_nonequil/output_active/*"
+        agents:
+          slurm_mem: 20GB
+
   - group: "Diagnostic EDMFX"
     steps:
 
@@ -724,7 +734,7 @@ steps:
         artifact_paths: "prognostic_edmfx_simpleplume_column/output_active/*"
         agents:
           slurm_mem: 20GB
-      
+
       - label: ":genie: Prognostic EDMFX Soares in a column"
         command: >
           julia --color=yes --project=.buildkite .buildkite/ci_driver.jl
@@ -787,6 +797,7 @@ steps:
         artifact_paths: "prognostic_edmfx_bomex_column_implicit/output_active/*"
         agents:
           slurm_mem: 20GB
+          slurm_constraint: icelake|cascadelake|skylake|epyc
 
       - label: ":genie: Prognostic EDMFX Dycoms RF01 in a column"
         command: >
@@ -805,7 +816,6 @@ steps:
         artifact_paths: "prognostic_edmfx_rico_column/output_active/*"
         agents:
           slurm_mem: 20GB
-        soft_fail: true
 
       - label: ":umbrella: Prognostic EDMFX TRMM in a column"
         command: >
@@ -815,7 +825,6 @@ steps:
         artifact_paths: "prognostic_edmfx_trmm_column/output_active/*"
         agents:
           slurm_mem: 20GB
-        soft_fail: true
 
       - label: ":genie: Prognostic EDMFX TRMM with 0-moment in a column"
         command: >
@@ -836,6 +845,15 @@ steps:
           slurm_mem: 20GB
         soft_fail: true
 
+      - label: ":genie: Prognostic EDMFX ERA5 Time Varying driven in a column"
+        command: >
+          julia --color=yes --project=.buildkite .buildkite/ci_driver.jl
+          --config_file $CONFIG_PATH/prognostic_edmfx_tv_era5driven_column.yml
+          --job_id prognostic_edmfx_tv_era5driven_column
+        artifact_paths: "prognostic_edmfx_tv_era5driven_column/output_active/*"
+        agents:
+          slurm_mem: 12GB
+
       - label: ":genie: Prognostic EDMFX Bomex in a box"
         command: >
           julia --color=yes --project=.buildkite .buildkite/ci_driver.jl
@@ -850,9 +868,14 @@ steps:
           julia --color=yes --project=.buildkite .buildkite/ci_driver.jl
           --config_file $CONFIG_PATH/prognostic_edmfx_aquaplanet.yml
           --job_id prognostic_edmfx_aquaplanet
+
+          julia --color=yes --project=.buildkite reproducibility_tests/test_mse.jl
+          --job_id prognostic_edmfx_aquaplanet
+          --out_dir prognostic_edmfx_aquaplanet/output_active
         artifact_paths: "prognostic_edmfx_aquaplanet/output_active/*"
         agents:
           slurm_mem: 20GB
+          slurm_constraint: icelake|cascadelake|skylake|epyc
 
   - group: "GPU"
     steps:
@@ -895,7 +918,7 @@ steps:
         depends_on:
           - "baroclinic_wave"
           - "baroclinic_wave_gpu"
-          
+
 
       - label: "GPU: baroclinic wave - 2 gpus"
         key: "baroclinic_wave_2gpu"
@@ -945,7 +968,7 @@ steps:
         agents:
           slurm_gpus: 1
           slurm_mem: 20GB
-          
+
       - label: "GPU: Diagnostic EDMFX aquaplanet"
         key: "diagnostic_edmfx_aquaplanet_gpu"
         command: >
@@ -962,7 +985,7 @@ steps:
       - label: "GPU: Prognostic EDMFX aquaplanet"
         command: >
           julia --color=yes --project=.buildkite .buildkite/ci_driver.jl
-          --config_file $CONFIG_PATH/prognostic_edmfx_aquaplanet.yml
+          --config_file $CONFIG_PATH/prognostic_edmfx_aquaplanet_gpu.yml
           --job_id prognostic_edmfx_aquaplanet_gpu
         artifact_paths: "prognostic_edmfx_aquaplanet_gpu/output_active/*"
         env:
@@ -1000,7 +1023,7 @@ steps:
           --job_id bm_default
         agents:
           slurm_mem: 24GB
-      
+
       - label: ":computer: Benchmark: GPU default"
         command: >
           julia --color=yes --project=.buildkite perf/benchmark.jl
@@ -1067,7 +1090,7 @@ steps:
         artifact_paths: "flame_default_1m/*"
         agents:
           slurm_mem: 24GB
-      
+
       - label: ":fire: Flame graph: diagnostics"
         command: >
           julia --color=yes --project=.buildkite perf/flame.jl
@@ -1094,7 +1117,7 @@ steps:
         artifact_paths: "flame_aquaplanet_progedmf/*"
         agents:
           slurm_mem: 32GB
-      
+
       - label: ":fire: Flame graph: diffusion"
         command: >
           julia --color=yes --project=.buildkite perf/flame.jl

.buildkite/pipeline.yml

@@ -6,7 +6,7 @@ println()
 println("julia -e 'using Pkg; Pkg.add(\"PkgDevTools\")'")
 println()
 println("Then, update the manifests with")
-println("using PkgDevTools; PkgDevToos.update_deps(\".\")")
+println("using PkgDevTools; PkgDevTools.update_deps(\".\")")
 println("in a Julia REPL.")
 println("See documentation to read more about this change")
 println("This file will be removed in future releases")

.dev/up_deps.jl

@@ -18,7 +18,7 @@ jobs:
   test:
     name: ci ${{ matrix.version }} - ${{ matrix.os }}
     runs-on: ${{ matrix.os }}
-    timeout-minutes: 45
+    timeout-minutes: ${{ matrix.timeout }}
     strategy:
       fail-fast: false
       matrix:
@@ -29,6 +29,11 @@ jobs:
           - ubuntu-latest
           - macOS-latest
           - windows-latest
+        include:
+          - version: '1.10'
+            timeout: 45
+          - version: '1.11'
+            timeout: 60
     steps:
       - uses: actions/checkout@v4
       - uses: julia-actions/setup-julia@v2

.github/workflows/ci.yml

@@ -7,6 +7,12 @@ git-tree-sha1 = "235a16e02c24a28b20244c0f2e7f149cc783166f"
 [era5_cloud]
 git-tree-sha1 = "8bad412ada94be95a2e11c794b499c49d746be50"
 
+[era5_hourly_atmos_raw]
+git-tree-sha1 = "8234def2ead82e385a330a48ed2f0c030e434065"
+
+[era5_hourly_atmos_processed]
+git-tree-sha1 = "a1a465e8d237d78bef1e6d346054da395787a9f9"
+
 [cfsite_gcm_forcing]
 git-tree-sha1 = "2f196bcac8950f8cb79ead41821f423d092b8951"
 lazy = true