AWS

Create data and WAL buckets (they can be the same bucket). Copy harness/aws/.env.example to harness/aws/.env and fill DATA_BUCKET and WAL_BUCKET. AWS_* is optional and depends on the credential provider.

Stream Stack

docker compose --env-file harness/aws/.env \
  -f harness/aws/docker-compose.native.yml \
  up -d --build

Durable Streams

docker compose --env-file harness/aws/.env \
  -f harness/aws/docker-compose.ds.yml \
  up -d --build

The node listens on 127.0.0.1:4437.

The following example has three nodes, on ports 4437-4439:

Stream Stack

docker compose --env-file harness/aws/.env \
  -f harness/aws/docker-compose.cluster.native.yml \
  up -d --build

Durable Streams

docker compose --env-file harness/aws/.env \
  -f harness/aws/docker-compose.cluster.ds.yml \
  up -d --build

Bare JVM (credentials from env / ~/.aws / IAM):

Stream Stack

java -jar frontend/native/server/target/native-server.jar \
  --topo harness/aws/topo.yaml --node-id 1 \
  --storage "0@s3://${DATA_BUCKET}?region=${AWS_REGION}" \
  --wal "0@s3://${WAL_BUCKET}?region=${AWS_REGION}"

Durable Streams

java -jar frontend/ds/server/target/ds-server.jar \
  --topo harness/aws/topo.yaml --node-id 1 \
  --storage "0@s3://${DATA_BUCKET}?region=${AWS_REGION}" \
  --wal "0@s3://${WAL_BUCKET}?region=${AWS_REGION}"

Topology

File	Storage
aws/topo.yaml	Single node
aws/topo.cluster.yaml	Multi-node (3 nodes)