(v2.2.0.0) Prerequisites : IO+HPC-slurm

Cluster

Software

Slurm from SchedMD

Permission

Cluster administrator access

Allow:

• Editing of slurm configuration files

• Restarting slurm

• root access to slurm head node and compute nodes

Network

VPC

Virtual networking resources

IPv4 CIDR block:

• Please specify a range other than 10.0.0.0/24 (Recommended)

• Subnet ID that Slurm compute nodes should be launched.

NAT gateways:

• Please create at least one

Security Group

The security group that is applied to cluster compute nodes.

Inbound Rules:

Outbound Rules:

Cluster Endpoint Access

AWS EC2 API server endpoint access

Either access to internet provided endpoints or EC2 VPC endpoint is required.

Compute

Node Groups

Allows EC2 instances to call AWS services on your behalf

Required Policies:

[AWS managed]

• AmazonEC2ContainerRegistryReadOnly → This allows read-only access to Amazon EC2 Container Registry repositories

• AmazonEKS_CNI_Policy → This provides the Amazon VPC CNI Add-on permissions it requires to modify the IP address configuration on your EKS worker nodes

• AmazonEKSWorkerNodePolicy → This allows Amazon EKS worker nodes to connect to Amazon EKS Clusters

• AmazonSSMManagedInstanceCore → This is to enable AWS Systems Manager service core functionality

Launch Templates

AMI

• AMI should not have AWS Marketplace product codes

• Slurmd should be disabled at boot in systemd

Add-ons