Deployment

The node is a single binary / container. Persist the database and checkpoint directories on a volume, and expose the admin, transfer and listener ports.

Docker

docker run -d --name pesitwizard \
  -p 8080:8080 -p 9081:9081 -p 5001:5001 \
  -e PESIT_API_KEY=secret \
  -e PESIT_DB=/data/pesitwizard.db \
  -v pesit-data:/data \
  ghcr.io/pesitwizard/pesitwizard:latest

Kubernetes

Deploy it as a normal Deployment + Service, with a PersistentVolumeClaim for /data and the built-in probes:

livenessProbe:
  httpGet: { path: /actuator/health/liveness, port: 8080 }
readinessProbe:
  httpGet: { path: /actuator/health/readiness, port: 8080 }

Scrape /metrics with Prometheus (no API key). For a highly-available deployment, run several replicas joined to a NATS cluster so configuration replicates and schedules are distributed; listener ports and connector credentials stay per-pod.

Configuration as code

Mount a PESIT_CONFIG YAML file (partners, virtual files, remote partners, listeners). It is applied at start-up and re-applied on SIGHUP, so a kubectl rollout restart or a config-map change can refresh configuration without losing the node's state.