Infrastructure

Infrastructure Reference

Quick reference for Odin HPC cluster infrastructure details.

Network Topology

graph TB
    Internet[Internet/VPN] --> JH[Jump Host<br/>jump.odin.navify.com]
    
    subgraph "Private Subnet"
        JH --> LN1[Login Node 1<br/>login1.odin.cluster.local]
        JH --> LN2[Login Node 2<br/>login2.odin.cluster.local]
        JH --> DML[Data Manager Linux<br/>data-manager-linux.odin.cluster.local]
        JH --> DMW[Data Manager Windows<br/>data-manager-windows.odin.cluster.local]
        
        LN1 --> HN[Head Node]
        LN2 --> HN
        HN --> CN[Compute Nodes]
    end

Instance Details

Jump Host

Property Value
DNS jump.odin.navify.com
Instance Type c6i.xlarge
vCPUs 4
Memory 8 GB
Purpose SSH bastion gateway

Login Nodes

Property Value
DNS login1.odin.cluster.local, login2.odin.cluster.local
Instance Type c6i.2xlarge
vCPUs 8
Memory 16 GB
Storage /mnt/odin, /mnt/qcs, /home, /mnt/shared
Purpose User sessions, job submission

Head Node

Property Value
Instance Type c6i.xlarge
vCPUs 4
Memory 4 GB
Purpose SLURM scheduler only

Warning: Do not SSH to headnode directly. Use login nodes.

Data Manager Linux

Property Value
DNS data-manager-linux.odin.cluster.local
Instance Type t3.medium
OS Ubuntu 22.04
Storage /mnt/odin, /mnt/qcs, /mnt/gxp, /mnt/ingest
Purpose Samba gateway, GxP data management

Data Manager Windows

Property Value
DNS data-manager-windows.odin.cluster.local
Instance Type t3.large
OS Windows Server 2025
Access RDP, SSH (PowerShell)
Purpose Windows data management

Compute Partitions

Partition Instance vCPUs Memory GPUs Max Nodes
cpu c7i.8xlarge 32 61 GB - 10
gpu-inferencing g5.8xlarge 32 122 GB 1× A10G 5
odin p5.48xlarge 192 1.9 TB 8× H100 2
albus p5.48xlarge 192 1.9 TB 8× H100 2
bali p5.48xlarge 192 1.9 TB 8× H100 2
genius p5.48xlarge 192 1.9 TB 8× H100 2

Storage Systems

Mount Type Size Available On
/mnt/odin FSx Lustre 12 TB All nodes, data managers
/mnt/qcs FSx Lustre 1.2 TB All nodes, data managers
/mnt/gxp FSx Lustre 1.2 TB Data managers only
/mnt/ingest FSx Lustre 9.6 TB Data managers only
/home EFS Elastic All nodes
/mnt/shared EBS Variable All nodes

DNS Zones

Zone Type Resolver
odin.navify.com Public Route53 Public DNS
odin.cluster.local Private Route53 VPC DNS (10.0.0.2)

Key AWS Resources

Terraform Outputs

cd odin/terraform

# Get all outputs
terraform output

# Specific outputs
terraform output jump_host_public_ip
terraform output odin_pcluster_headnode
terraform output data_manager_linux_private_ip
terraform output data_manager_windows_private_ip

AWS Secrets

Secret Purpose
odin/samba-users/{username} Samba/Windows passwords
/odin/data-manager-windows/initial-password Windows admin password
odin-db-connection Database credentials

Key Parameters

Parameter Purpose
/odin/data-manager-windows/initial-password Windows initial password

GitHub Actions Workflows

Workflow Trigger Purpose
update-users Manual Sync users to all instances
terraform-plan PR Plan infrastructure changes
terraform-apply Merge to main Apply infrastructure changes