​

Distributed Deployment

​

Recommended Local Cluster

./scripts/start_cluster.sh start --clean --build

• backend nodes on 3000, 3001, 3002
• dashboard UI on 3100

​

Cluster Management Endpoints

• POST /v1/cluster/initialize
• POST /v1/cluster/join
• DELETE /v1/cluster/nodes/:node_id

​

Join Payload

{
  "node_id": 2,
  "address": "127.0.0.1:5002"
}

​

Example Initialization

curl -s -X POST http://127.0.0.1:3000/v1/cluster/initialize \
  -H "Authorization: Bearer $TOKEN"

​

Example Join

curl -s -X POST http://127.0.0.1:3000/v1/cluster/join \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "node_id": 2,
    "address": "127.0.0.1:5002"
  }'

​

Operational Notes

• In single-node mode, join and leave operate on the local raft group.
• In sharded mode, cluster operations fan out across shard groups.
• Leader forwarding is built into write paths, but operational cluster control still expects a healthy Raft topology.