Node Build

Deploy an Akash RPC node using one of three methods: CLI build, Helm chart (Kubernetes), or Omnibus (on Akash Network).

Why run an Akash node?

• Required for validators
• Recommended for providers (lower latency)
• Best practice for production dApps (no reliance on public nodes)

Deployment Methods

CLI Build

Manual installation on a Linux server.

Time: 15-30 minutes (setup) + sync time

Requirements:

• Ubuntu 24.04 LTS
• 4 CPU cores
• 8 GB RAM
• 100 GB SSD storage (minimum), 1 TB recommended
• Root access

Use when:

• You want full control over the node configuration
• Running on bare metal or VPS
• Learning node operations from scratch

Helm Chart

Deploy to an existing Kubernetes cluster via Helm.

Time: 5-10 minutes

Requirements:

• Existing Kubernetes cluster
• Helm 4.0+
• kubectl access
• 100 GB storage (minimum), 1 TB recommended

Use when:

• You already have a Kubernetes cluster
• You want quick deployment with blockchain snapshots
• You need to manage multiple nodes

Omnibus

Deploy on Akash Network using a pre-built SDL.

Time: 5-10 minutes (deployment) + 20-30 minutes (sync via hourly snapshot)

Requirements:

• Akash wallet with ~2 AKT
• Akash Console or CLI access

Use when:

• You want to run a node on Akash’s network
• Testing or learning Akash deployments
• You don’t have dedicated infrastructure

After Deployment

Once your node is running:

1. Verify sync status - Check if catching_up: false
2. Check peer connections - Ensure connected to seed/peer nodes
3. Monitor block height - Compare with Mintscan
4. Enable services - Configure RPC/API if needed

For validators, see Running a Validator.

Questions? Join #validators on Discord