retrying a failed request on an already overloaded system made the overloaded situation worse.
Rebuilding the Segment Leaderboards Infrastructure — Part 1: Background
Jeff Pollard
4181

Hey Jeff, great post. I’m curious if your previous retry logic implemented randomness (otherwise known as jitter - https://www.awsarchitectureblog.com/2015/03/backoff.html). Not that it would have fixed the problem overloading you in the first place, just curious with respect to retry implementations.

Thanks,

Ben

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.