Comment 11 for bug 1572719

I think this is medium impact at best - it's very much like many other slow resource exhaustion from ill behaved clients or bad requests we've squashed in the past. Generally these are mitigated in public environments with monitoring and alerting - although I'm sure they can be annoying!

Once the eventlet issue is public - and especially following a release - this should of course become public!

I'd prefer to eat the upstream version rev than to try and implement a hacky work-around in our tree.

... IMHO