[do not merge] C4a blueprints

examples/README.md

@@ -16,6 +16,8 @@ md_toc github examples/README.md | sed -e "s/\s-\s/ * /"
 * [Instructions](#instructions)
   * [(Optional) Setting up a remote terraform state](#optional-setting-up-a-remote-terraform-state)
 * [Blueprint Descriptions](#blueprint-descriptions)
+  * [c4a-vm.yaml](#c4a-vmyaml-) ![core-badge]
+  * [hpc-slurm-c4a.yaml](#hpc-slurm-c4ayaml-) ![core-badge]
   * [hpc-slurm.yaml](#hpc-slurmyaml-) ![core-badge]
   * [hpc-enterprise-slurm.yaml](#hpc-enterprise-slurmyaml-) ![core-badge]
   * [hpc-slurm-static.yaml](#hpc-slurm-staticyaml-) ![core-badge]
@@ -29,7 +31,7 @@ md_toc github examples/README.md | sed -e "s/\s-\s/ * /"
   * [serverless-batch-mpi.yaml](#serverless-batch-mpiyaml-) ![core-badge]
   * [pfs-lustre.yaml](#pfs-lustreyaml-) ![core-badge] ![deprecated-badge]
   * [pfs-managed-lustre-vms.yaml](#pfs-managed-lustre-vmsyaml-) ![core-badge]
-  * [gke-managed-lustre.yaml](#gke-managed-lustreyaml-) ![core-badge]  
+  * [gke-managed-lustre.yaml](#gke-managed-lustreyaml-) ![core-badge]
   * [ps-slurm.yaml](#ps-slurmyaml--) ![core-badge] ![experimental-badge]
   * [cae-slurm.yaml](#cae-slurmyaml-) ![core-badge]
   * [hpc-build-slurm-image.yaml](#hpc-build-slurm-imageyaml--) ![community-badge] ![experimental-badge]
@@ -159,6 +161,50 @@ Toolkit team, partners, etc.) and are labeled with the community badge
 Blueprints that are still in development and less stable are also labeled with
 the experimental badge (![experimental-badge]).
 
+### [c4a-vm.yaml] ![core-badge]
+
+The [c4a-vm.yaml] blueprint creates a small, two-node cluster of C4A (ARM64) VMs. It includes a Filestore instance mounted to `/home` and uses Hyperdisk for the boot disks.
+
+To deploy this blueprint:
+
+```bash
+gcluster deploy examples/c4a-vm.yaml \
+    -v project_id=<YOUR-PROJECT-ID> \
+    -v deployment_name=c4a-vm
+```
+
+#### Quota Requirements for c4a-vm.yaml
+
+For this example the following is needed in the selected region:
+
+* Cloud Filestore API: Basic HDD (Standard) capacity (GB): **1,024 GB**
+* Compute Engine API: Hyperdisk Balanced (GB): **~100 GB** (50 GB/node)
+* Compute Engine API: C4A CPUs: **144** (2 nodes * 72 vCPUs)
+
+[c4a-vm.yaml]: ./c4a-vm.yaml
+
+### [hpc-slurm-c4a.yaml] ![core-badge]
+
+This blueprint creates a small Slurm cluster using C4A (ARM64) instances. It provisions a `c4a-standard-4` for the controller and login nodes, and a dynamic partition of `c4a-highcpu-72` compute nodes. All nodes use Hyperdisk Balanced for their boot disks. It also includes a 2.5 TiB Basic SSD Filestore instance for `/home`. This blueprint is a good starting point for running ARM-based workloads on a Slurm cluster.
+
+To deploy this blueprint:
+
+```bash
+gcluster deploy examples/hpc-slurm-c4a.yaml \
+    -v project_id=<YOUR-PROJECT-ID> \
+    -v deployment_name=slurm-c4a
+```
+
+#### Quota Requirements for hpc-slurm-c4a.yaml
+
+For this example the following is needed in the selected region:
+
+* Cloud Filestore API: Basic SSD capacity (GB): **2,560 GB**
+* Compute Engine API: Hyperdisk Balanced (GB): **~200 GB** (50 GB/node for 2 compute nodes + 50 GB for controller boot disk + 50 GB for controller state disk + 50 GB for login)
+* Compute Engine API: C4A CPUs: **152** (4 for controller + 4 for login + 2 nodes * 72 vCPUs)
+
+[hpc-slurm-c4a.yaml]: ./hpc-slurm-c4a.yaml
+
 ### [hpc-slurm.yaml] ![core-badge]
 
 Creates a basic auto-scaling Slurm cluster with mostly default settings. The
@@ -786,7 +832,7 @@ providing a high-performance file system for demanding workloads.
      volumes:
      - name: lustre-volume
        persistentVolumeClaim:
-         claimName: $(vars.lustre_instance_id)-pvc # Matches the PVC name  
+         claimName: $(vars.lustre_instance_id)-pvc # Matches the PVC name
    ```
 
    Note: This is just an example job using busybox image.
@@ -1172,7 +1218,7 @@ to the cluster using `kubectl` and will run on the specified node pool.
     1. The output of the `./gcluster deploy` on CLI includes a `kubectl create` command to create the job.
 
        ```sh
-       kubectl create -f <job-yaml-path> 
+       kubectl create -f <job-yaml-path>
        ```
 
        This command creates a job that uses busybox image and prints `Hello World`. This result can be viewed by looking at the pod logs.

examples/c4a-vm.yaml

@@ -0,0 +1,72 @@
+# Copyright 2025 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.
+
+---
+
+blueprint_name: c4a-vm
+
+vars:
+  project_id: ## Set GCP Project ID Here ##
+  deployment_name: c4a-vm
+  region: us-central1
+  zone: us-central1-a
+  hostname_prefix: $(vars.deployment_name)
+  base_network_name: $(vars.deployment_name)
+  instance_image:
+    family: rocky-linux-8-optimized-gcp-arm64
+    project: rocky-linux-cloud
+
+# Documentation for each of the modules used below can be found at
+# https://github.com/GoogleCloudPlatform/hpc-toolkit/blob/main/modules/README.md
+
+deployment_groups:
+- group: primary
+  modules:
+
+  # Source is an embedded module, denoted by "modules/*" without ./, ../, /
+  # as a prefix. To refer to a local module, prefix with ./, ../ or /
+
+  - id: cluster-net-0
+    source: modules/network/vpc
+    settings:
+      network_name: $(vars.base_network_name)-net
+
+  - id: homefs
+    source: modules/file-system/filestore
+    use: [cluster-net-0]
+    settings:
+      local_mount: /home
+    outputs:
+    - network_storage
+
+  - id: c4a_startup
+    source: modules/scripts/startup-script
+    settings:
+      configure_ssh_host_patterns:
+      - $(vars.hostname_prefix)-*
+
+  - id: c4a-vms
+    source: modules/compute/vm-instance
+    use: [c4a_startup, homefs, cluster-net-0]
+    settings:
+      machine_type: c4a-standard-72
+      instance_count: 2
+      disk_type: hyperdisk-balanced
+      bandwidth_tier: tier_1_enabled
+
+  - id: wait-for-vms
+    source: community/modules/scripts/wait-for-startup
+    settings:
+      instance_names: $(c4a-vms.name)
+      timeout: 7200

examples/hpc-slurm-c4a.yaml

@@ -0,0 +1,101 @@
+# Copyright 2025 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.
+---
+blueprint_name: slurm-c4a
+vars:
+  project_id: ## Set GCP Project ID Here ##
+  deployment_name: slurm-c4a
+  region: us-central1
+  zone: us-central1-a
+  instance_image:
+    family: slurm-gcp-6-9-ubuntu-2204-lts-arm64
+    project: schedmd-slurm-public
+  disk_type: hyperdisk-balanced
+
+# Documentation for each of the modules used below can be found at
+# https://github.com/GoogleCloudPlatform/hpc-toolkit/blob/main/modules/README.md
+deployment_groups:
+- group: primary
+  modules:
+
+  # Source is an embedded module, denoted by "modules/*" without ./, ../, /
+  # as a prefix. To refer to a local module, prefix with ./, ../ or /
+
+  - id: c4a-slurm-net-0
+    source: modules/network/vpc
+
+  - id: homefs
+    source: modules/file-system/filestore
+    use: [c4a-slurm-net-0]
+    settings:
+      filestore_tier: BASIC_SSD
+      size_gb: 2560
+      filestore_share_name: homeshare
+      local_mount: /home
+
+  - id: c4a_startup
+    source: modules/scripts/startup-script
+    settings:
+      local_ssd_filesystem:
+        fs_type: ext4
+        mountpoint: /mnt/lssd
+        permissions: "1777"
+
+  - id: c4a_nodeset
+    source: community/modules/compute/schedmd-slurm-gcp-v6-nodeset
+    use: [c4a_startup, c4a-slurm-net-0]
+    settings:
+      # Unattended upgrades are disabled to prevent automatic daily updates which may lead to potential instability.
+      # For more details, see https://cloud.google.com/compute/docs/instances/create-hpc-vm#disable_automatic_updates
+      allow_automatic_updates: false
+      bandwidth_tier: tier_1_enabled
+      machine_type: c4a-highcpu-72
+      node_count_static: 0
+      node_count_dynamic_max: 2
+      enable_placement: true
+
+  - id: c4a_partition
+    source: community/modules/compute/schedmd-slurm-gcp-v6-partition
+    use: [c4a_nodeset]
+    settings:
+      exclusive: false
+      partition_name: c4a
+      is_default: true
+      partition_conf:
+        ResumeTimeout: 900
+        SuspendTimeout: 600
+
+  - id: login_startup
+    source: modules/scripts/startup-script
+    settings:
+      configure_ssh_host_patterns:
+      - $(vars.deployment_name)-*
+
+  - id: slurm_login
+    source: community/modules/scheduler/schedmd-slurm-gcp-v6-login
+    use: [c4a-slurm-net-0]
+    settings:
+      machine_type: c4a-standard-4
+      enable_login_public_ips: true
+
+  - id: slurm_controller
+    source: community/modules/scheduler/schedmd-slurm-gcp-v6-controller
+    use: [c4a-slurm-net-0, c4a_partition, slurm_login, homefs]
+    settings:
+      machine_type: c4a-standard-4
+      controller_state_disk:
+        type: hyperdisk-balanced
+        size: 50
+      enable_controller_public_ips: false
+      login_startup_script: $(login_startup.startup_script)