3 ms·
> ” The reason that we haven’t been doing any significant work on this HAProxy stack is that we’re moving towards Envoy Proxy ” I truly don’t understand the cy
by nikanj 6y ago
> ” The reason that we haven’t been doing any significant work on this HAProxy stack is that we’re moving towards Envoy Proxy ”
I truly don’t understand the cycle of churn. Once most edge cases and bugs of Haproxy have been found, the right decision is not migrating to completely unknown territory again. No project is a silver bullet, and changing stacks after you find the bugs makes for a terrible return on your bug-hunting investments.
- bastawhiz 6y agoAnd if there is no path forward to resolve the bugs? How many rough edges and bugs can you tolerate before a system is no longer worth investing in, especially when another newer system does not suffer from any of those bugs? To say there is no tipping point where the cost of continuing to use a system which no longer fits your needs exceeds the cost of migrating to a system which does (and will continue to) fit your needs ignores some amount of common sense.
- brown9-2 6y agoI think if you read this as they’re moving to Envoy because of this incident then you’ve misread. But it also sounds like Envoy and HAProxy have fundamentally different approaches to service discovery: https://twitter.com/mattklein123/status/1277729102271676416?s=21 https://twitter.com/mattklein123/status/1277729102271676416?... https://twitter.com/mattklein123/status/1278114953497436161?s=21 https://twitter.com/mattklein123/status/1278114953497436161?...