AlloyDB Omni Samples

This repository provides best-practice technical examples and manifests for deploying and operating the AlloyDB Omni Operator on Kubernetes.

Here you will find templates and guidance to automate your database life cycle operations, whether you are managing clusters individually or managing fleets using enterprise GitOps tools.

Documentation Guides

We have organized the examples in this repository into comprehensive functional guides. Please explore the operations relevant to your use case:

• Cluster Management: Deploy basic to full configuration primary databases, setup read pools, and configure PostgreSQL parameters.
• Backup and Restore: Schedule automated backup plans, push to Google Cloud Storage (GCS), trigger manual backups, and perform point-in-time recovery (PITR) clones.
• Replication and Disaster Recovery: Setup Upstream/Downstream data replication across clusters, and orchestrate graceful switchovers or forced failovers.
• Helm and GitOps Deployments: Discover how to use our Helm charts and ArgoCD configurations to simplify management of the operator and database workloads.
• Monitoring and Extensibility: Leverage our Grafana dashboards for metrics visibility and see examples of using custom mutating webhooks to dynamically enforce topology constraints.

Repository Structure

• samples/: Raw Kubernetes YAML custom resources for all the life cycle capabilities discussed in the guides.
• k8s/helm/: Helm Charts for deploying the AlloyDB Omni Operator and configurable Clusters.
• crd/: The raw Custom Resource Definitions (CRDs) required by the operator.
• monitoring-dashboards/: Extracted dashboard configurations for integration with observability stacks.
• webhooks/: Go-based sample mutating webhook servers mapped to database operations.
• experimental/: Community-driven experimental tools, applications and observability stacks.

Quickstart

If you already have the operator running, you can deploy a sample minimal database in seconds by applying our base configuration:

kubectl apply -f samples/v1_dbcluster.yaml

Note: You must edit the secret configuration if deploying into production.