Merge pull request #2924 from GoogleCloudPlatform/release-candidate

Release v1.38.0
GoogleCloudPlatform · Aug 15, 2024 · 1e38ce0 · 1e38ce0
2 parents 229803f + 0ef79d3
commit 1e38ce0
Show file tree

Hide file tree

Showing 324 changed files with 10,602 additions and 1,061 deletions.
diff --git a/.github/workflows/dependency-review.yml b/.github/workflows/dependency-review.yml
@@ -26,7 +26,7 @@ permissions:
 
 jobs:
   dependency-review:
-    if: github.repository == 'GoogleCloudPlatform/hpc-toolkit'
+    if: github.repository == 'GoogleCloudPlatform/cluster-toolkit'
     runs-on: ubuntu-latest
     steps:
     - name: 'Checkout Repository'

diff --git a/.github/workflows/pr-label-validation.yml b/.github/workflows/pr-label-validation.yml
@@ -32,7 +32,7 @@ on:
 
 jobs:
   pr-label-validation:
-    if: github.repository == 'GoogleCloudPlatform/hpc-toolkit'
+    if: github.repository == 'GoogleCloudPlatform/cluster-toolkit'
     runs-on: ubuntu-latest
     permissions:
       pull-requests: read

diff --git a/.github/workflows/pr-precommit.yml b/.github/workflows/pr-precommit.yml
@@ -38,6 +38,10 @@ jobs:
         python-version: '3.10'
         check-latest: true
         cache: 'pip'
+    - run: >
+        pip install
+        -r community/modules/scheduler/schedmd-slurm-gcp-v6-controller/modules/slurm_files/scripts/requirements.txt
+        -r community/modules/scheduler/schedmd-slurm-gcp-v6-controller/modules/slurm_files/scripts/tests/requirements.txt
     - uses: actions/setup-go@v5
       with:
         go-version: '1.22'

diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml
@@ -74,6 +74,12 @@ repos:
     files: 'tools/cloud-build/daily-tests/builds/.*\.yaml'
     pass_filenames: false
     require_serial: true
+  - id: pytest-check
+    name: pytest-check
+    entry: pytest
+    language: system
+    types: [python]
+    pass_filenames: false
 
 - repo: https://github.com/dnephin/pre-commit-golang
   rev: v0.5.1

diff --git a/.pytest.ini b/.pytest.ini
@@ -0,0 +1,17 @@
+# Copyright 2024 Google LLC
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+[pytest]
+filterwarnings = ignore::DeprecationWarning
+testpaths = tests community/modules/scheduler/schedmd-slurm-gcp-v6-controller/modules/slurm_files/scripts/tests
diff --git a/README.md b/README.md
@@ -33,8 +33,8 @@ If a self directed path is preferred, you can use the following commands to
 build the `gcluster` binary:
 
 ```shell
-git clone https://github.com/GoogleCloudPlatform/hpc-toolkit
-cd hpc-toolkit
+git clone https://github.com/GoogleCloudPlatform/cluster-toolkit
+cd cluster-toolkit
 make
 ./gcluster --version
 ./gcluster --help

diff --git a/cmd/root.go b/cmd/root.go
@@ -22,6 +22,7 @@ import (
 	"fmt"
 	"hpc-toolkit/pkg/config"
 	"hpc-toolkit/pkg/logging"
+	"hpc-toolkit/pkg/shell"
 	"os"
 	"os/exec"
 	"path/filepath"
@@ -52,7 +53,7 @@ HPC deployments on the Google Cloud Platform.`,
 				logging.Fatal("cmd.Help function failed: %s", err)
 			}
 		},
-		Version:     "v1.37.2",
+		Version:     "v1.38.0",
 		Annotations: annotation,
 	}
 )
@@ -79,13 +80,21 @@ func Execute() error {
 		if len(GitBranch) == 0 {
 			GitBranch = "detached HEAD"
 		}
+
 		annotation["version"] = GitTagVersion
 		annotation["branch"] = GitBranch
 		annotation["commitInfo"] = GitCommitInfo
-		rootCmd.SetVersionTemplate(`gcluster version {{index .Annotations "version"}}
+		tmpl := `gcluster version {{index .Annotations "version"}}
 Built from '{{index .Annotations "branch"}}' branch.
 Commit info: {{index .Annotations "commitInfo"}}
-`)
+`
+		tfVersion, _ := shell.TfVersion()
+		if tfVersion != "" {
+			annotation["tfVersion"] = tfVersion
+			tmpl += `Terraform version: {{index .Annotations "tfVersion"}}
+`
+		}
+		rootCmd.SetVersionTemplate(tmpl)
 	}
 
 	return rootCmd.Execute()

diff --git a/community/examples/AMD/hpc-amd-slurm-v5-legacy.yaml b/community/examples/AMD/hpc-amd-slurm-v5-legacy.yaml
@@ -171,7 +171,7 @@ 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-5-11-hpc-centos-7
+        family: slurm-gcp-5-12-hpc-centos-7
         project: schedmd-slurm-public
 
   - id: low_cost_node_group

diff --git a/community/examples/AMD/hpc-amd-slurm.yaml b/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-6-5-hpc-rocky-linux-8
+        family: slurm-gcp-6-6-hpc-rocky-linux-8
         project: schedmd-slurm-public
 
   - id: low_cost_nodeset
@@ -179,6 +179,7 @@ deployment_groups:
       node_count_dynamic_max: 10
       bandwidth_tier: gvnic_enabled
       enable_placement: false
+      allow_automatic_updates: false
 
   - id: low_cost_partition
     source: community/modules/compute/schedmd-slurm-gcp-v6-partition
@@ -194,6 +195,7 @@ deployment_groups:
       node_count_dynamic_max: 50
       bandwidth_tier: gvnic_enabled
       enable_placement: true
+      allow_automatic_updates: false
 
   # Because is_default is set to true, jobs will run on this partition unless an
   # alternative partition is specified using, for example, "srun -p lowcost"

diff --git a/community/examples/client-google-cloud-storage.yaml b/community/examples/client-google-cloud-storage.yaml
@@ -57,6 +57,7 @@ deployment_groups:
     settings:
       name_prefix: workstation
       machine_type: e2-standard-2
+      allow_automatic_updates: false
 
   - id: wait
     source: community/modules/scripts/wait-for-startup

diff --git a/community/examples/hpc-build-slurm-image.yaml b/community/examples/hpc-build-slurm-image.yaml
@@ -104,6 +104,7 @@ deployment_groups:
     settings:
       machine_type: n2d-standard-2
       instance_image: $(vars.built_instance_image)
+      allow_automatic_updates: false
 
   - id: debug_partition
     source: community/modules/compute/schedmd-slurm-gcp-v6-partition

diff --git a/community/examples/hpc-slurm-chromedesktop-v5-legacy.yaml b/community/examples/hpc-slurm-chromedesktop-v5-legacy.yaml
@@ -22,10 +22,10 @@ vars:
   region: us-central1
   zone: us-central1-c
   instance_image_crd:
-    family: slurm-gcp-5-11-debian-11
+    family: slurm-gcp-5-12-debian-11
     project: schedmd-slurm-public
   instance_image:
-    family: slurm-gcp-5-11-hpc-centos-7
+    family: slurm-gcp-5-12-hpc-centos-7
     project: schedmd-slurm-public
 
 # Documentation for each of the modules used below can be found at

diff --git a/community/examples/hpc-slurm-gromacs.yaml b/community/examples/hpc-slurm-gromacs.yaml
@@ -97,6 +97,7 @@ deployment_groups:
     settings:
       node_count_dynamic_max: 20
       bandwidth_tier: gvnic_enabled
+      allow_automatic_updates: false
 
   - id: compute_partition
     source: community/modules/compute/schedmd-slurm-gcp-v6-partition

diff --git a/community/examples/hpc-slurm-local-ssd-v5-legacy.yaml b/community/examples/hpc-slurm-local-ssd-v5-legacy.yaml
@@ -37,6 +37,19 @@ deployment_groups:
     settings:
       local_mount: /home
 
+  - id: startup
+    source: modules/scripts/startup-script
+    settings:
+      # When shutting down a VM with local SSD disks, we strongly recommend the
+      # automatic migration of data following these instructions:
+      # https://cloud.google.com/compute/docs/disks/local-ssd#stop_instance
+      # Failure to do will result in VMs that lose data and do not automatically
+      # mount local SSD filesystems
+      local_ssd_filesystem:
+        fs_type: ext4
+        mountpoint: /mnt/localssd
+        permissions: "1777" # must quote numeric filesystem permissions!
+
   - id: compute_node_group
     source: community/modules/compute/schedmd-slurm-gcp-v5-node-group
     settings:
@@ -66,39 +79,11 @@ deployment_groups:
     - network1
     - homefs
     - compute_node_group
+    - startup
     settings:
       is_default: true
       partition_name: ssdcomp
       region: us-central1
-      startup_script: |
-        #!/bin/bash
-        set -e -o pipefail
-
-        # this script assumes it is running on a RedHat-derivative OS
-        yum install -y mdadm
-
-        RAID_DEVICE=/dev/md0
-        DST_MNT=/mnt/localssd
-        DISK_LABEL=LOCALSSD
-        OPTIONS=discard,defaults
-
-        # if mount is successful, do nothing
-        if mount --source LABEL="$DISK_LABEL" --target="$DST_MNT" -o "$OPTIONS"; then
-                exit 0
-        fi
-
-        # Create new RAID, format ext4 and mount
-        # TODO: handle case of zero or 1 local SSD disk
-        # TODO: handle case when /dev/md0 exists but was not mountable
-        DEVICES=`nvme list | grep nvme_ | grep -v nvme_card-pd | awk '{print $1}' | paste -sd ' '`
-        NB_DEVICES=`nvme list | grep nvme_ | grep -v nvme_card-pd | awk '{print $1}' | wc -l`
-        mdadm --create "$RAID_DEVICE" --level=0 --raid-devices=$NB_DEVICES $DEVICES
-        mkfs.ext4 -F "$RAID_DEVICE"
-        tune2fs "$RAID_DEVICE" -r 131072
-        e2label "$RAID_DEVICE" "$DISK_LABEL"
-        mkdir -p "$DST_MNT"
-        mount --source LABEL="$DISK_LABEL" --target="$DST_MNT" -o "$OPTIONS"
-        chmod 1777 "$DST_MNT"
 
   - id: slurm_controller
     source: community/modules/scheduler/schedmd-slurm-gcp-v5-controller

diff --git a/community/examples/hpc-slurm-local-ssd.yaml b/community/examples/hpc-slurm-local-ssd.yaml
@@ -37,9 +37,22 @@ deployment_groups:
     settings:
       local_mount: /home
 
+  - id: startup
+    source: modules/scripts/startup-script
+    settings:
+      # When shutting down a VM with local SSD disks, we strongly recommend the
+      # automatic migration of data following these instructions:
+      # https://cloud.google.com/compute/docs/disks/local-ssd#stop_instance
+      # Failure to do will result in VMs that lose data and do not automatically
+      # mount local SSD filesystems
+      local_ssd_filesystem:
+        fs_type: ext4
+        mountpoint: /mnt/localssd
+        permissions: "1777" # must quote numeric filesystem permissions!
+
   - id: nodeset
     source: community/modules/compute/schedmd-slurm-gcp-v6-nodeset
-    use: [network]
+    use: [network, startup]
     settings:
       additional_disks:
       - device_name: test-disk-1
@@ -60,35 +73,7 @@ deployment_groups:
       machine_type: c2-standard-4
       node_count_dynamic_max: 5
       node_count_static: 0
-      startup_script: |
-        #!/bin/bash
-        set -e -o pipefail
-
-        # this script assumes it is running on a RedHat-derivative OS
-        yum install -y mdadm
-
-        RAID_DEVICE=/dev/md0
-        DST_MNT=/mnt/localssd
-        DISK_LABEL=LOCALSSD
-        OPTIONS=discard,defaults
-
-        # if mount is successful, do nothing
-        if mount --source LABEL="$DISK_LABEL" --target="$DST_MNT" -o "$OPTIONS"; then
-                exit 0
-        fi
-
-        # Create new RAID, format ext4 and mount
-        # TODO: handle case of zero or 1 local SSD disk
-        # TODO: handle case when /dev/md0 exists but was not mountable
-        DEVICES=`nvme list | grep nvme_ | grep -v nvme_card-pd | awk '{print $1}' | paste -sd ' '`
-        NB_DEVICES=`nvme list | grep nvme_ | grep -v nvme_card-pd | awk '{print $1}' | wc -l`
-        mdadm --create "$RAID_DEVICE" --level=0 --raid-devices=$NB_DEVICES $DEVICES
-        mkfs.ext4 -F "$RAID_DEVICE"
-        tune2fs "$RAID_DEVICE" -r 131072
-        e2label "$RAID_DEVICE" "$DISK_LABEL"
-        mkdir -p "$DST_MNT"
-        mount --source LABEL="$DISK_LABEL" --target="$DST_MNT" -o "$OPTIONS"
-        chmod 1777 "$DST_MNT"
+      allow_automatic_updates: false
 
   - id: partition
     source: community/modules/compute/schedmd-slurm-gcp-v6-partition

diff --git a/community/examples/hpc-slurm-ramble-gromacs.yaml b/community/examples/hpc-slurm-ramble-gromacs.yaml
@@ -118,6 +118,7 @@ deployment_groups:
     settings:
       node_count_dynamic_max: 20
       bandwidth_tier: gvnic_enabled
+      allow_automatic_updates: false
 
   - id: compute_partition
     source: community/modules/compute/schedmd-slurm-gcp-v6-partition

diff --git a/community/examples/hpc-slurm-sharedvpc.yaml b/community/examples/hpc-slurm-sharedvpc.yaml
@@ -62,6 +62,7 @@ deployment_groups:
       node_count_dynamic_max: 4
       machine_type: n2-standard-2
       enable_placement: false  # the default is: true
+      allow_automatic_updates: false
 
   - id: debug_partition
     source: community/modules/compute/schedmd-slurm-gcp-v6-partition
@@ -77,6 +78,7 @@ deployment_groups:
     settings:
       node_count_dynamic_max: 20
       bandwidth_tier: gvnic_enabled
+      allow_automatic_updates: false
 
   - id: compute_partition
     source: community/modules/compute/schedmd-slurm-gcp-v6-partition

diff --git a/community/examples/hpc-slurm-ubuntu2004-v5-legacy.yaml b/community/examples/hpc-slurm-ubuntu2004-v5-legacy.yaml
@@ -24,7 +24,7 @@ vars:
   instance_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-5-11-ubuntu-2004-lts
+    family: slurm-gcp-5-12-ubuntu-2004-lts
     project: schedmd-slurm-public
   instance_image_custom: true
 

diff --git a/community/examples/hpc-slurm-ubuntu2004.yaml b/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-6-5-ubuntu-2004-lts
+    family: slurm-gcp-6-6-ubuntu-2004-lts
     project: schedmd-slurm-public
   instance_image_custom: true
 

diff --git a/community/examples/hpc-slurm6-apptainer.yaml b/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-6-4-hpc-rocky-linux-8
+      source_image_family: slurm-gcp-6-6-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)
@@ -78,6 +78,7 @@ deployment_groups:
       instance_image: $(vars.custom_image)
       instance_image_custom: true
       bandwidth_tier: gvnic_enabled
+      allow_automatic_updates: false
 
   - id: compute_partition
     source: community/modules/compute/schedmd-slurm-gcp-v6-partition

diff --git a/community/examples/hpc-slurm6-tpu-maxtext.yaml b/community/examples/hpc-slurm6-tpu-maxtext.yaml
@@ -100,6 +100,7 @@ deployment_groups:
       name: ns2
       node_count_dynamic_max: 20
       bandwidth_tier: gvnic_enabled
+      allow_automatic_updates: false
 
   - id: compute_partition
     source: community/modules/compute/schedmd-slurm-gcp-v6-partition

diff --git a/community/examples/htc-htcondor.yaml b/community/examples/htc-htcondor.yaml
@@ -104,6 +104,7 @@ deployment_groups:
       name_prefix: grp1
       instance_image: $(vars.new_image)
       min_idle: 2
+      allow_automatic_updates: false
 
   - id: htcondor_execute_point_spot
     source: community/modules/compute/htcondor-execute-point
@@ -117,6 +118,7 @@ deployment_groups:
       name_prefix: spot
       instance_image: $(vars.new_image)
       spot: true
+      allow_automatic_updates: false
 
   - id: htcondor_access
     source: community/modules/scheduler/htcondor-access-point