Cleaning up full instance disk space
Written by
Updated at July 30, 2024
A GitLab instance can run out of disk space. This is indicated by the 500 Internal Server Error
HTTP status code. In which case you will not be able to access the GitLab instance. To clean up the disk and recover access to the GitLab instance, contact technical support
To reduce the probability of running out of disk space in future:
- Clean up the instance disk space manually at regular intervals.
- Configure disk space cleanup policies.
Clean up the instance disk space
- Manually review
all Docker images and their tags in GitLab Container Registry. - Delete
outdated images and tags.
Configure disk space cleanup policies
You can use any of the following options:
- Set the expiration time for build artifacts
at the instance level. Its default value is 30 days. - Create and configure a tag cleanup policy
at the level of individual projects utilizing GitLab Container Registry. - If you are using a Yandex Container Registry integration, create and configure a Docker image cleanup policy for Yandex Container Registry.