CROWDB / DEPLOYMENT

Deployment guides

Bring up the storage layers in dependency order.

Development preview · bare-metal deployment is not production-ready

1. KV cluster

Launch KV nodes and publish Group 0 identity from a sealed bootstrap input.

Start with KV →

2. Chunk layer

Configure DiskIO, DiskDB and ChunkDB after Group 0 is ready.

Configure storage →

3. Data access

Attach S3 or Iceberg access to a ready storage cluster.

Configure access →

Docker preview

Use the independent single-node evaluation image for disposable data.

Open quick start →

These guides describe development-stage operation and health checks. Do not treat the bare-metal path as production-ready. Group 0 owns cluster topology; local launch registries own only process commands and credentials are resolved locally.