Why is my stack still using an old runner image after updating it?

Last updated: August 6, 2025

Why is my stack still using an old runner image after updating it?

If you've updated your runner image but notice that new runs are still using the old image, there are a few key things to check:

1. Check your config.yml file

First, verify that your config.yml file doesn't contain a reference to the old runner image. Even if you've updated the image elsewhere, the configuration file can take precedence.

More details here.

Tips for Faster Image Updates

If you need to prioritize runs with the new image:

  • For private workers, you can manually prioritize specific runs either through the UI or by adjusting your push policy

  • While forcing a worker refresh through autoscaling groups will terminate old workers, verify that no configuration is causing new workers to pull the old image

Note: Even after forcing a worker refresh, if you continue to see the old image being used, double-check your configuration files as they may be the source of the old image reference.