Enterprise installation
Choose Kubernetes for a production deployment. Local source checkout and Compose are useful for evaluation and development; see Setup. Enterprise images and charts require the registry credentials supplied with your delivery. They are separate artefacts from the public Community images.
Prerequisites
- A supported Kubernetes cluster, Helm 3.8+ and a configured kubectl context.
- PostgreSQL and a database/user/DSN prepared for nxs-anomaly. The example uses an external database; PostgreSQL 17 is the reference deployment image.
- The exact release tag, Enterprise chart OCI reference and registry account supplied with the delivery. Use one matching version for API, worker and UI.
- An image pull Secret provisioned by the platform in the application namespace.
- Bash and a secure location for the application environment file.
Kubernetes with external PostgreSQL
Set EE_CHART_REF to the supplied nxs-anomaly-enterprise chart reference, RELEASE_VERSION to its version and REGISTRY_HOST to its registry. Authenticate Helm without putting the password in shell history:
read -r -p 'Registry username: ' REGISTRY_USER
read -r -s -p 'Registry password: ' REGISTRY_PASSWORD; printf '\n'
printf '%s' "$REGISTRY_PASSWORD" | helm registry login "$REGISTRY_HOST" \
--username "$REGISTRY_USER" --password-stdin
unset REGISTRY_PASSWORDCreate namespace nxs-anomaly, arrange its image pull Secret named registry-credentials, and prepare application.env with permissions 0600. It must contain NXS_ANOMALY_DB_DSN, bootstrap admin username/password and any provider credentials. Obtain DSN/TLS options from your database administrator; do not use a production password from an example.
kubectl create namespace nxs-anomaly
kubectl -n nxs-anomaly create secret generic nxs-anomaly-env \
--from-env-file=application.env
helm upgrade --install nxs-anomaly "$EE_CHART_REF" \
--version "$RELEASE_VERSION" --namespace nxs-anomaly \
--set existingSecret.enabled=true --set-string existingSecret.name=nxs-anomaly-env \
--set-string 'imagePullSecrets[0].name=registry-credentials' \
--wait --timeout 5m
kubectl -n nxs-anomaly port-forward service/nxs-anomaly-frontend 3100:8080Open http://localhost:3100 and sign in using the bootstrap credentials. Remove the bootstrap password variables from the Secret after verifying normal administrative access: keeping them resets that account's password on startup. The chart disables bundled PostgreSQL by default; the DSN above is required. Configure production ingress, certificates, backups, outbound transport and readiness checks before accepting incidents.
Local evaluation and development
Use the private source checkout provided with the delivery, not the GitHub CE checkout. Copy .env.example to .env, fill in the required database/bootstrap secrets and run docker compose up --build. Compose uses development defaults; its named PostgreSQL volume survives normal down, while down -v deletes data. The detailed component table, health checks and local Go commands are in Setup. Build tools must match go.mod; frontend requires Node22.12+.
For bare metal, use the same API/worker/frontend image versions or a build from the delivered source. Configure PostgreSQL, a same-origin reverse proxy and managed service startup as described in Deployment. Do not run the scheduler in every HTTP replica: use API --no-scheduler and dedicated workers.
Enable optional Enterprise components
Installation alone does not enable SSO or analytical data services. Configure OIDC through API authentication; configure Kafka/ClickHouse and the consumer through Analytics pipeline. Review High availability, Backup and restore, Security and Configuration before production rollout.