Create a hostile environment, perhaps a kubernetes cluster, where you can stress-test your deployment. Verify things like security settings, do load testing, restart pods often, etc.
- openapi (optional)
- deployment
- service
- docker inspect an image or a running container if possible,
docker inspect <image-name>
- Check that the deployment has a good security context set
- Call all endpoints with weird stuff
- slowloris
- Restart pods often
- Have a stressed client deployment that gets angry if it can't access the service