clusterdatabase-as-a-servicedata-infrastructuregolangkubernetesmanaged-servicesoperatorpostgrespostgres-operatorpostgresql
acb88d9b93
In Kubernetes, there is tight coupling of API group, CRD and Custom Resources managed by operators. CRDs are cluster resource and unique per cluster. CRD operatorconfigurations.acid.zalan.do defines the operator configurations for postgres operator. CRD postgresqls.acid.zalan.do defines minimum and maximum supported postgresql versions. This leads to issues in a multi-tenant k8s cluster, an example provided below. In Namespace X, user X deploys web tier application which uses zalando postgres-operator with max version PG 15. In Namespace Y, user Y deploys web tier application which also uses zalando postgres-operator but needs PG 17 runs into an issue as CRD postgresqls.acid.zalan.do supports only PG 15. This happens because CRD is cluster scope and all namespace scoped applications should adhere to it. With large k8s clusters, postgresql being a very popular database, zalando postgres operator being defacto way to manage databases, there is a need to keep this configurable so applications can manage the API group of CRDs. Take API group for postgres operator as an env variable, assuming that the relevant CRDs with API groups are installed. |
||
|---|---|---|
| .github | ||
| charts | ||
| cmd | ||
| docker | ||
| docs | ||
| e2e | ||
| hack | ||
| kubectl-pg | ||
| logical-backup | ||
| manifests | ||
| mocks | ||
| pkg | ||
| ui | ||
| .flake8 | ||
| .gitignore | ||
| .golangci.yml | ||
| .zappr.yaml | ||
| CODEOWNERS | ||
| CONTRIBUTING.md | ||
| LICENSE | ||
| MAINTAINERS | ||
| Makefile | ||
| README.md | ||
| SECURITY.md | ||
| build-ci.sh | ||
| delivery.yaml | ||
| go.mod | ||
| go.sum | ||
| mkdocs.yml | ||
| run_operator_locally.sh | ||
README.md
Postgres Operator
The Postgres Operator delivers an easy to run highly-available PostgreSQL clusters on Kubernetes (K8s) powered by Patroni. It is configured only through Postgres manifests (CRDs) to ease integration into automated CI/CD pipelines with no access to Kubernetes API directly, promoting infrastructure as code vs manual operations.
Operator features
- Rolling updates on Postgres cluster changes, incl. quick minor version updates
- Live volume resize without pod restarts (AWS EBS, PVC)
- Database connection pooling with PGBouncer
- Support fast in place major version upgrade. Supports global upgrade of all clusters.
- Restore and cloning Postgres clusters on AWS, GCS and Azure
- Additionally logical backups to S3 or GCS bucket can be configured
- Standby cluster from S3 or GCS WAL archive
- Configurable for non-cloud environments
- Basic credential and user management on K8s, eases application deployments
- Support for custom TLS certificates
- UI to create and edit Postgres cluster manifests
- Compatible with OpenShift
PostgreSQL features
- Supports PostgreSQL 17, starting from 13+
- Streaming replication cluster via Patroni
- Point-In-Time-Recovery with pg_basebackup / WAL-E via Spilo
- Preload libraries: bg_mon, pg_stat_statements, pgextwlist, pg_auth_mon
- Incl. popular Postgres extensions such as decoderbufs, hypopg, pg_cron, pg_partman, pg_stat_kcache, pgq, pgvector, plpgsql_check, postgis, set_user and timescaledb
The Postgres Operator has been developed at Zalando and is being used in production for over five years.
Supported Postgres & K8s versions
| Release | Postgres versions | K8s versions | Golang |
|---|---|---|---|
| v1.14.0 | 13 → 17 | 1.27+ | 1.23.4 |
| v1.13.0 | 12 → 16 | 1.27+ | 1.22.5 |
| v1.12.0 | 11 → 16 | 1.27+ | 1.22.3 |
| v1.11.0 | 11 → 16 | 1.27+ | 1.21.7 |
| v1.10.1 | 10 → 15 | 1.21+ | 1.19.8 |
| v1.9.0 | 10 → 15 | 1.21+ | 1.18.9 |
Getting started
For a quick first impression follow the instructions of this tutorial.
Supported setups of Postgres and Applications
Documentation
There is a browser-friendly version of this documentation at postgres-operator.readthedocs.io