Years ago, when I was playing with AWS in a course on building cloud-hosted services, it was well-known that all the AWS management was hosted out of a single zone, and there were several days we had to cancel class because us-east-1 had an outage, so while technically all our VMs hosted out of other AZs were extant, all our attempts to manage our VMs via the web UI or API were timing or erroring out.
I understand this is long-since resolved (I haven't tried building a service on Amazon in a couple years, so this isn't personal experience), but centralized failure modes in decentralized systems can persist longer than you might expect.
(Work for Google, not on Cloud or anything related to this outage that I'm aware of, I have no knowledge other than reading the linked outage page.)