Backend to schedule Bakta jobs on a Kubernetes cluster.
The Bakta-Web backend implements a simple job scheduling system for the Bakta-Web UI. It requires an S3 compatible object storage, a Kubernetes cluster with ArgoWorkflows to run the jobs. Jobs can be submitted via an API. The API is a simple REST API. The corresponding repositories can be found here:
- S3-compatible object storage
- Kubernetes-Cluster
- Argo Workflows with a pre-configured WorkflowTemplate named
bakta-job-{BAKTA_VERSTION}
The Container accepts the following settings via env-vars (or .env file):
SOCKET_ADDR=127.0.0.1:8080
ARGO_TOKEN=token
ARGO_URL=https://argo.example.com
ARGO_NAMESPACE=argo
S3_ACCESS_KEY=access_key
S3_SECRET_KEY=secret_key
S3_BUCKET=bucket
S3_ENDPOINT=https://s3.example.com
BAKTA_VERSION=0.1.0
DATABASE_VERSION=0.1.0
BACKEND_VERSION=0.2.0