Thanks, yes protocol-aware recovery was the context. Pretty sure I first heard it described in Joran's QCon London 2023 talk here: https://youtu.be/_jfOk4L7CiY?

Thanks, yes protocol-aware recovery was the context. Pretty sure I first heard it described in Joran's QCon London 2023 talk here: https://youtu.be/_jfOk4L7CiY?t=1460 https://youtu.be/_jfOk4L7CiY?t=1460 > If you want your distributed database to maximise availability, how your local storage engine recovers from storage faults in the write-ahead log needs to be properly integrated with the global consensus protocol.