Project reference ↗

A distributed database has several components and an ongoing lifecycle of configuration changes, scaling and upgrades. TiDB Operator provides Kubernetes controllers for managing TiDB deployments from declared resources, giving platform teams an operating interface for that database system. It is useful when the team has chosen TiDB’s distributed SQL model and wants Kubernetes-based management. Installing the controller does not create a complete production database: storage, failure placement, access and tested recovery still need to be configured.

Chart ownership

PingCAP's stable installation guide documents the pingcap/tidb-operator chart from charts.pingcap.com and a separate CRD installation step. Its development documentation describes a newer OCI distribution. Select one released operator generation and follow its matching documentation; do not combine stable values and CRDs with development installation examples.

Before adoption

Installing the controller does not create a complete production database. Define the TiDB cluster resources, storage classes, failure domains and backups separately. Pin operator, database and CRD versions and review their compatibility. Test failover and a restore into an isolated environment, and verify application behavior with realistic transactions. Restrict controller permissions and database access. Plan upgrades for the operator and the managed database as related but distinct changes.

Sources & further reading

  1. TiDB Operator stable installation
  2. TiDB Kubernetes quick start
  3. TiDB Operator development documentation

Spotted something that needs another look?

Help improve this page →