Secrets creep into dashboards more often than coffee spills on keyboards. Everyone knows it’s wrong, yet there they are—API keys, database credentials, maybe even a private token—stuffed into Looker connections. The fix feels simple until credentials need rotation or compliance knocks. That’s where HashiCorp Vault Looker integration comes in.
Vault is the trusted vault for storing secrets, encrypting data, and managing dynamic credentials. Looker, Google’s data platform, needs secure access to databases and APIs. Connect them correctly, and you get automated credential handoffs without exposing passwords to developers or dashboards. Connect them poorly, and you get 3 a.m. alerts about expired tokens and failed queries.
The magic is in the handshake. Vault issues short-lived, scoped credentials. Looker then fetches them on demand, usually through an external service account or plugin that authenticates via OIDC or an intermediary like AWS IAM. Instead of embedding secrets, you fetch what you need, when you need it, from Vault’s API. Configuration effort pays off immediately—no more static credentials, no messy rotation schedules, just identity-aware access.
Most teams start by linking Vault’s identity management to Looker’s service account model. Map roles to policies in Vault that match Looker’s data connections. Then grant read permissions for specific secrets only. Keep the TTLs short. Rotate automatically. Done right, you’ll never have to manually update a credential file again.
If something fails mid-integration, the issue is almost always around permissions. Vault policies that are too strict cause Looker connection errors. Policies that are too broad ruin your compliance story. Test each path with simulation commands before deploying, and watch your audit trail like a hawk.