AWS ParallelCluster 3.4.0 でコンピュートノードのマルチ AZ 起動をサポートしました

  SlurmQueues:
    # ------ Compute 1 ------
    - Name: queue-1
      ComputeResources:
        - Name: multiaz-queue
          Instances:
            - InstanceType: c6i.large
          MinCount: 0
          MaxCount: 10
          DisableSimultaneousMultithreading: true
      ComputeSettings:
        LocalStorage:
          RootVolume:
            Size: 35
            Encrypted: true
            VolumeType: gp3
            Iops: 3000
            Throughput: 125
      CapacityType: SPOT
      AllocationStrategy: lowest-price
      Networking:
        SubnetIds:
          - subnet-035be95eeaa091603
          - subnet-0ba7369f9caba6f93

Region: ap-northeast-1
Image:
  Os: ubuntu2004
Tags:
  - Key: Name
    Value: MultiAZ-Cluster
# ----------------------------------------------------------------
# Head Node Settings
# ----------------------------------------------------------------
HeadNode:
  InstanceType: t3.micro
  Networking:
    ElasticIp: false
    SubnetId: subnet-035be95eeaa091603
  Ssh:
    KeyName: sandbox-key
  LocalStorage:
    RootVolume:
      Size: 35
      Encrypted: true
      VolumeType: gp3
      Iops: 3000
      Throughput: 125
  Iam:
    AdditionalIamPolicies:
      - Policy: arn:aws:iam::aws:policy/AmazonSSMManagedInstanceCore
    S3Access:
      - BucketName: hpc-dev-postinstall-files
        EnableWriteAccess: false
# ----------------------------------------------------------------
# Compute Node Settings
# ----------------------------------------------------------------
Scheduling:
  Scheduler: slurm
  SlurmSettings:
    ScaledownIdletime: 5
  SlurmQueues:
    # ------ Compute 1 ------
    - Name: queue-1
      ComputeResources:
        - Name: multiaz-queue
          Instances:
            - InstanceType: c6i.large
          MinCount: 0
          MaxCount: 10
          DisableSimultaneousMultithreading: true
      ComputeSettings:
        LocalStorage:
          RootVolume:
            Size: 35
            Encrypted: true
            VolumeType: gp3
            Iops: 3000
            Throughput: 125
      CapacityType: SPOT
      AllocationStrategy: lowest-price
      Networking:
        SubnetIds:
          - subnet-035be95eeaa091603
          - subnet-0ba7369f9caba6f93
        # PlacementGroup:
        #  Enabled: true
      # CustomActions:
      #   OnNodeConfigured:
      #     Script: s3://hpc-dev-postinstall-files/sample-ubuntu-docker/postinstall.sh
      Iam:
        AdditionalIamPolicies:
          - Policy: arn:aws:iam::aws:policy/AmazonSSMManagedInstanceCore
        S3Access:
          - BucketName: hpc-dev-postinstall-files
            EnableWriteAccess: false

# ----------------------------------------------------------------
# Shared Storage Settings
# ----------------------------------------------------------------
# SharedStorage:
#   - MountDir: /mnt/efs-1zone
#     Name: efs-1zone
#     StorageType: Efs
#     EfsSettings:
#       FileSystemId: fs-0f1158ade79354809
# ----------------------------------------------------------------
#  Other Settings
# ----------------------------------------------------------------
Monitoring:
  Logs:
    CloudWatch:
      Enabled: true
      RetentionInDays: 30
      DeletionPolicy: "Delete"
  Dashboards:
    CloudWatch:
      Enabled: false

{
  "configurationValidationErrors": [
    {
      "level": "ERROR",
      "type": "MultiAzPlacementGroupValidator",
      "message": "You have enabled PlacementGroups for the 'multiaz-queue' Compute Resource on the 'queue-1' queue. PlacementGroups are not supported across Availability zones. Either remove the PlacementGroup configuration to use multiple subnets on the queue or specify only one subnet to use a PlacementGroup for compute resources."
    },
    {
      "level": "ERROR",
      "type": "EfsIdValidator",
      "message": "Cluster has subnets located in different availability zones but EFS (fs-0f1158ade79354809) uses OneZone EFS storage class which works within a single Availability Zone. Please use subnets located in one Availability Zone or use a standard storage class EFS."
    },
    {
      "level": "INFO",
      "type": "MultiAzRootVolumeValidator",
      "message": "Your configuration for Queues 'queue-1' includes multiple subnets different from where HeadNode is located. Accessing a shared storage from different AZs can lead to increased storage network latency and inter-AZ data transfer costs."
    }
  ],
  "message": "Invalid cluster configuration."
}