merge develop

.github/workflows/pr-label-validation.yml

@@ -28,11 +28,13 @@ on:
     - ready_for_review
     - unlocked
     branches:
+    - main
     - develop
+    - release-candidate
 
 jobs:
   pr-label-validation:
-    if: github.repository == 'GoogleCloudPlatform/cluster-toolkit'
+    if: github.repository == 'GoogleCloudPlatform/cluster-toolkit' && github.event.pull_request.draft == false
     runs-on: ubuntu-latest
     permissions:
       pull-requests: read

.github/workflows/pr-precommit.yml

@@ -19,9 +19,8 @@ name: 'Use pre-commit to validate Pull Request'
 on:
   pull_request:
     types:
-    - edited
     - opened
-    - labeled
+    - reopened
     - synchronize
     branches:
     - main

cmd/create.go

@@ -125,9 +125,27 @@ func expandOrDie(path string) (config.Blueprint, *config.YamlCtx) {
 	// Expand the blueprint
 	checkErr(bp.Expand(), ctx)
 	validateMaybeDie(bp, *ctx)
+	v5DeprecationWarning(bp)
+
 	return bp, ctx
 }
 
+// TODO: Remove this warning when v5 deprecation is complete
+func v5DeprecationWarning(bp config.Blueprint) {
+	alreadyContainsV5 := false
+	bp.WalkModulesSafe(func(mp config.ModulePath, m *config.Module) {
+		if strings.Contains(m.Source, "schedmd-slurm-gcp-v5-controller") && !alreadyContainsV5 {
+			logging.Info(boldYellow(
+				"We have been supporting slurm-gcp v5 since July 2022 and are now deprecating it, as we've launched slurm-gcp v6 in June 2024. \n" +
+					"Toolkit blueprints using Slurm-gcp v5 will be marked “deprecated” starting October 2024 and slurm-gcp v6 will be the default deployment. \n" +
+					"However we won't begin removing slurm-gcp v5 blueprints until January 6, 2025. Beginning on January 6, 2025, the Cluster Toolkit team will cease their support for Slurm-gcp v5. \n" +
+					"While this will not directly or immediately impact running clusters, we recommend replacing any v5 clusters with Slurm-gcp v6.",
+			))
+			alreadyContainsV5 = true // This is to avoid the logging message showing repeatedly for multiple v5 controllers
+		}
+	})
+}
+
 // TODO: move to expand.go
 func validateMaybeDie(bp config.Blueprint, ctx config.YamlCtx) {
 	err := validators.Execute(bp)

cmd/root.go

@@ -53,7 +53,7 @@ HPC deployments on the Google Cloud Platform.`,
 				logging.Fatal("cmd.Help function failed: %s", err)
 			}
 		},
-		Version:     "v1.39.0",
+		Version:     "v1.41.0",
 		Annotations: annotation,
 	}
 )

community/examples/AMD/hpc-amd-slurm.yaml

@@ -168,7 +168,7 @@ deployment_groups:
         # these images must match the images used by Slurm modules below because
         # we are building OpenMPI with PMI support in libraries contained in
         # Slurm installation
-        family: slurm-gcp-6-6-hpc-rocky-linux-8
+        family: slurm-gcp-6-8-hpc-rocky-linux-8
         project: schedmd-slurm-public
 
   - id: low_cost_nodeset

community/examples/hpc-build-slurm-image.yaml

@@ -23,7 +23,7 @@ vars:
   image_build_machine_type: n2d-standard-16
   build_from_image_family: hpc-rocky-linux-8
   build_from_image_project: cloud-hpc-image-public
-  build_from_git_ref: 6.7.0
+  build_from_git_ref: 6.8.2
   built_image_family: my-custom-slurm
   built_instance_image:
     family: $(vars.built_image_family)

community/examples/hpc-slurm-ramble-gromacs.yaml

@@ -31,7 +31,6 @@ deployment_groups:
   modules:
   # Source is an embedded module, denoted by "modules/*" without ./, ../, /
   # as a prefix. To refer to a local module, prefix with ./, ../ or /
-  # Example - ./modules/network/vpc
   - id: network
     source: modules/network/vpc
 

community/examples/hpc-slurm-ubuntu2004-v5-legacy.yaml

@@ -34,7 +34,6 @@ deployment_groups:
   modules:
   # Source is an embedded module, denoted by "modules/*" without ./, ../, /
   # as a prefix. To refer to a local module, prefix with ./, ../ or /
-  # Example - ./modules/network/vpc
   - id: network1
     source: modules/network/vpc
 

community/examples/hpc-slurm-ubuntu2004.yaml

@@ -24,7 +24,7 @@ vars:
   slurm_image:
     # Please refer to the following link for the latest images:
     # https://github.com/GoogleCloudPlatform/slurm-gcp/blob/master/docs/images.md#supported-operating-systems
-    family: slurm-gcp-6-6-ubuntu-2004-lts
+    family: slurm-gcp-6-8-ubuntu-2004-lts
     project: schedmd-slurm-public
   instance_image_custom: true
 
@@ -33,7 +33,6 @@ deployment_groups:
   modules:
   # Source is an embedded module, denoted by "modules/*" without ./, ../, /
   # as a prefix. To refer to a local module, prefix with ./, ../ or /
-  # Example - ./modules/network/vpc
   - id: network1
     source: modules/network/vpc
 

community/examples/hpc-slurm6-apptainer.yaml

@@ -60,7 +60,7 @@ deployment_groups:
     settings:
       source_image_project_id: [schedmd-slurm-public]
       # see latest in https://github.com/GoogleCloudPlatform/slurm-gcp/blob/master/docs/images.md#published-image-family
-      source_image_family: slurm-gcp-6-6-hpc-rocky-linux-8
+      source_image_family: slurm-gcp-6-8-hpc-rocky-linux-8
       # You can find size of source image by using following command
       # gcloud compute images describe-from-family <source_image_family> --project schedmd-slurm-public
       disk_size: $(vars.disk_size)

community/examples/hpc-slurm6-tpu.yaml

@@ -49,6 +49,7 @@ deployment_groups:
     use: [tpu_nodeset]
     settings:
       partition_name: tpu
+      resume_timeout: 600
 
   - id: slurm_login
     source: community/modules/scheduler/schedmd-slurm-gcp-v6-login

community/examples/htc-slurm-v5-legacy.yaml

@@ -42,7 +42,6 @@ deployment_groups:
   modules:
   # Source is an embedded module, denoted by "modules/*" without ./, ../, /
   # as a prefix. To refer to a local or community module, prefix with ./, ../ or /
-  # Example - ./modules/network/pre-existing-vpc
   - id: network1
     source: modules/network/vpc
 

community/examples/htc-slurm.yaml

@@ -42,7 +42,6 @@ deployment_groups:
   modules:
   # Source is an embedded module, denoted by "modules/*" without ./, ../, /
   # as a prefix. To refer to a local or community module, prefix with ./, ../ or /
-  # Example - ./modules/network/pre-existing-vpc
   - id: network
     source: modules/network/vpc