> Before cloud-native applications, MTTR meant one thing – Mean Time to Repair.
I remember some things in colos that needed to periodically be restarted. That's definitely "recover" and not "repair."
I remember one company that would detect that one of its non-cloud-workers was getting overwhelmed, automatically remove it from the work pool, then re-insert it when it could breathe again. That's automated recovery.
The biggest difference in pre-cloud and pre-container was that the mechanisms to handle this were all bespoke.
