cluster.yaml

Region: <REGION>
Image:
  Os: ubuntu2004
HeadNode:
  InstanceType: m6i.large
  Networking:
    SubnetId: <HEADNODE_SUBNET>
    AdditionalSecurityGroups:
      - <ADDITIONAL_HEAD_NODE_SG>
  Ssh:
    KeyName: <KEY_NAME>
    AllowedIps: <ALLOWED_IPS>
  Iam:
    AdditionalIamPolicies:
      - Policy: arn:aws:iam::aws:policy/AmazonSSMManagedInstanceCore
      - Policy: arn:aws:iam::aws:policy/CloudWatchLogsReadOnlyAccess
      - Policy: <ADDITIONAL_HEAD_NODE_POLICY>
    S3Access:
      - BucketName: <BUCKET_NAME>
        EnableWriteAccess: false
  CustomActions:
    OnNodeConfigured:
      Script: s3://<BUCKET_NAME>/post_install.sh
Scheduling:
  Scheduler: slurm
  SlurmSettings:
    QueueUpdateStrategy: DRAIN
  SlurmQueues:
  - Name: compute
    ComputeResources:
    - Name: cpu
      InstanceType: c6i.large
      MinCount: 0
      MaxCount: 5
    Networking:
      SubnetIds:
        - <COMPUTE_SUBNET>
    Iam:
      AdditionalIamPolicies:
        - Policy: arn:aws:iam::aws:policy/AmazonSSMManagedInstanceCore
        - Policy: <ADDITIONAL_COMPUTE_NODE_POLICY>
      S3Access:
        - BucketName: <BUCKET_NAME>
          EnableWriteAccess: false
    CustomActions:
      OnNodeConfigured:
        Script: s3://<BUCKET_NAME>/post_install.sh
  - Name: gpu
    ComputeResources:
    - Name: gpu
      InstanceType: g4dn.xlarge
      MinCount: 0
      MaxCount: 5
    Networking:
      SubnetIds:
        - <Compute_Subnet>
    Iam:
      AdditionalIamPolicies:
        - Policy: arn:aws:iam::aws:policy/AmazonSSMManagedInstanceCore
        - Policy: <ADDITIONAL_COMPUTE_NODE_POLICY>
      S3Access:
        - BucketName: <BUCKET_NAME>
          EnableWriteAccess: false
    CustomActions:
      OnNodeConfigured:
        Script: s3://<BUCKET_NAME>/post_install.sh
SharedStorage:
  - MountDir: /data_lustre
    Name: fsx-lustre-data
    StorageType: FsxLustre
    FsxLustreSettings:
      StorageCapacity: 1200
      DeploymentType: SCRATCH_2