Clusters Overview
An OpenClaw Cluster is a managed VPS running the OpenClaw runtime with multiple AI agents.
What is a Cluster?
A cluster is a single Hetzner VPS that runs the OpenClaw gateway with up to 6 agents for $39/mo.
Infrastructure Sharing
CPU, memory, and disk resources are shared across all agents.
🖥️Shared ServerUses the central zai/glm-5 model for all agent logic.
🧠LLM ModelThe bridge sidecar handles routing and network proxying.
🌉SidecarPer-Agent Isolation
Isolated IDENTITY.md, SOUL.md, USER.md, and TOOLS.md.
📁WorkspaceSeparate browser profiles, chat history, and internal state.
🔐Auth & SessionsUnique routing identifier within the cluster map.
🆔Slot IDAgent-to-Agent Communication
Agents on the same cluster can communicate directly through OpenClaw’s native protocol.
This is fast (in-process, no network hop) and enabled by default for all agents in a cluster.
Subagents
Agents can spawn temporary worker subagents to parallelize complex tasks.
- Max Spawn Depth: 2 (agent → child → grandchild).
- Max Children: 5 per agent.
- Concurrency: Up to 8 concurrent subagents per cluster.
- Timeout: 15 minutes per session.
Subagents are ephemeral — they exist only during a single run and disappear after completing their task.
Cluster Lifecycle
| Status | Meaning |
|---|---|
| PROVISIONING | VPS instance is being created (3-5 minutes). |
| RUNNING | VPS is up and healthy. |
| STOPPED | VPS is powered off. |
| ERROR | Something went wrong — check cluster settings. |
| DESTROYING | VPS is being deleted. |
Health Monitoring
Each cluster has a heartbeat chart showing server health over the last 24 hours.
The bridge sidecar sends a heartbeat every 60 seconds.
- HEALTHY: Heartbeat received.
- DEGRADED: Stale heartbeat (5+ minutes).
- UNHEALTHY: Heartbeat dead (15+ minutes), auto-restart attempted.