Hacker News Reader

Written by a guy who’s livelihood is supporting AWS customers.

I have no strong opinions on the topic, but the author sounds like an asshole going for the impression that confidence == competence.

View on HN

I've been using a few of the mentioned providers (RedisLabs, Elastic Cloud to name a few). I hoped because these companies have control over the software they support, who else would know better how to operate them? It is not just about providing the best throughput for buck, but to avoid JVM memory pressure hangups or providing details why the query response time went up and the CPU utilisation stayed the same? Why there is a 2-minute gap in the metrics? (Instance restarted, but who knows?) Many frustrating issues that seems to be so easy to solve and they would have run into these if they have to be forced to use their own product. AWS has its own issues, but at least more transparent when it comes to understanding hickups.

Reply on news.ycombinator.com