STREAM STACKDurable streams on object storage

Storage, metadata, and protocol as decoupled layers.

Get started

Docs

Infinite streams

Create as many as you need. Partitions are now streams, no topic tax.

Cheap storage

Cost effective with S3-compatible object storage for WAL and data.

High throughput

Battle-tested S3Stream storage backend. Transparent benchmark suite.

High availability

N nodes, one cluster. JRaft replicated metadata.

Open source

Apache 2.0 licensed. No hidden capabilities.

Many protocols

The Stream Stack API, plus durable-streams compatibility on the same engine.

Create streams as you go.

Streams can match the logical granularity of each use case,
rather than collecting every record of a kind into one topic.

Unlimited streams

Stream data and the WAL live on S3-compatible object storage, so create as many as you need with no topic tax.

$ docker compose --env-file harness/local/.env \
    -f harness/local/docker-compose.minio.yml \
    -f harness/local/docker-compose.ds.yml \
    up -d --build
✔ Node listening on 127.0.0.1:4437

$ java -jar cli/target/streamstack.jar ds bench \
    --endpoint http://127.0.0.1:4437 \
    -b 1024 -n 256 -w 32 -d 20
✔ Write: 69.45 MiB/s, 71117 records/s

Blazing fast

Battle-tested S3Stream storage backend for high throughput. Transparent, reproducible benchmarks.

Multi-node

N nodes, one cluster. JRaft replicated metadata for high availability and fault tolerance.

$ docker compose --env-file harness/local/.env \
    -f harness/local/docker-compose.minio.yml \
    -f harness/local/docker-compose.cluster.ds.yml \
    up -d --build
✔ Nodes on 127.0.0.1:4437-4439