Description
We build services, potentially used by millions of people. And despite all the testing we do, some interactions with these services will not work out the way we hope. Wouldn’t it be great to reconstruct what let to a problem and analyze if the problem occurred at other times?