Interesting insights into how Kinesis works under the hood. And an interesting discussion of a failure mode it seems like a company the size and maturity of AWS should have protected against (or at least detected sooner).

Looks like we were due for our yearly us-east-1 outage that takes a bunch of global services with it, impacting people in other regions.

Wonder if it's just the usual us-east-1 massive problems we see every so often, or if it's related to upcoming reinvent announcements.