This post looks at the difference between getting Authentik back itself, fixing access to an application tied to it, and getting into infrastructure when Authentik is completely down. I also walk through where I keep local break-glass access and why I would avoid having one shared emergency credential. Also, we look at why it is so important to test these accounts just like we test our backups.
Curious how others handle emergency access in their self-hosted environments without having just another single point of failure.