A data orchestration runtime, in Go. Workflows, a queue, a scheduler and a UI in one binary.
2.4K
A data orchestration runtime, in Go.
Declarative transformation, workflow orchestration, a persistent queue, a scheduler and an operational interface — in one binary. Every step runs as its own Kubernetes pod, with its own image.
| tag | base | role |
|---|---|---|
:<version> | distroless | The API and the UI. It runs nothing, so it has no shell. |
:<version>-worker | alpine | scheduler, publish, backfill. It has a shell, because a workflow's run: steps need one. |
The split is not fussiness. The worker runs arbitrary commands out of the client's YAML and the API does not, so giving the API a shell would widen the surface of the network-exposed process by the component that needs it least.
Both are built linux/amd64 and linux/arm64 from the same binary.
docker run --rm areteacademy/brevis:latest version
A full stack — Postgres, the API, the scheduler and a worker — is one compose file:
curl -O https://raw.githubusercontent.com/AreteAcademy/brevis/master/examples/quickstart/docker-compose.yml
docker compose up -d
open http://localhost:8080
MIT licensed.
Content type
Image
Digest
sha256:cd48b9856…
Size
7.4 MB
Last updated
2 days ago
docker pull areteacademy/brevis