Summary
Add a Helm chart that deploys a single-replica rsigma daemon on Kubernetes with ConfigMap-mounted rules, ServiceMonitor scraping, probes on the existing health endpoint, optional PVC for SQLite state, and NATS credentials via Secrets.
Motivation
The Docker image already ships, but operators still hand-author manifests. A chart is the production deployment story for single-replica streaming detection and the baseline for a later multi-replica chart.
Proposed approach
- Chart resources: Deployment, ConfigMap for rules, Service, ServiceMonitor, PDB, optional PVC for state.
- Wire readiness/liveness to the daemon HTTP health endpoint.
- NATS URL and TLS material via Secrets; document single-replica plus optional NATS assumptions.
- Values for image, resources, rule mounts, and sink/source config overlays.
Out of scope
- Multi-replica / StatefulSet clustering (separate follow-on).
- A full Kubernetes operator with CRDs (separate project).
Tasks
References
- Docker image and hardening notes in the root README
- Daemon health and metrics endpoints
- Existing
docker run documentation
Summary
Add a Helm chart that deploys a single-replica rsigma daemon on Kubernetes with ConfigMap-mounted rules, ServiceMonitor scraping, probes on the existing health endpoint, optional PVC for SQLite state, and NATS credentials via Secrets.
Motivation
The Docker image already ships, but operators still hand-author manifests. A chart is the production deployment story for single-replica streaming detection and the baseline for a later multi-replica chart.
Proposed approach
Out of scope
Tasks
deploy/helm/rsigma)References
docker rundocumentation