> For the complete documentation index, see [llms.txt](https://developer.harness.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://developer.harness.io/continuous-delivery/troubleshooting-and-resources/armory/general/setting-an-sql-cleanup-job-for-clouddriver.md).

# Setting an SQL Cleanup Job for Clouddriver

### Introduction <a href="#introduction" id="introduction"></a>

Spinnaker allows for the easy configuration of a cleanup job for the Clouddriver MySQL backend if it is enabled. This can be helpful to remove orphaned entries and maintain a reasonable size of the SQL backend. Enabling this setting causes clouddriver-caching's clean up agent to periodically purge old clusters and accounts. OSS recommends setting to true when using the Kubernetes provider.

### Prerequisites <a href="#prerequisites" id="prerequisites"></a>

N/A

### Instructions <a href="#instructions" id="instructions"></a>

**In Halyard:**

This would be added to the `clouddriver-local.yaml` file in the hal config profiles directory e.g. (\~/.hal/default/profiles/) `sql.unknown-agent-cleanup-agent.enabled: true` Alternatively, Users can manually go into the MYSQL Database and remove the offending entries.

**In Operator:**

This can be set at **spec.spinnakerConfig.profiles.clouddriver** spec: spinnakerConfig: profiles: clouddriver: sql.unknown-agent-cleanup-agent.enabled: true Alternatively, Users can manually go into the MYSQL Database and remove the offending entries.