1 This is a story of how we adapted our architecture over time to accomodate growth.
2 The response time of such a queue depends on all prior requests, so if we drastically increase the average work time of one request the average resp..
3 So we stuck with the model of a single concurrency per application server process and ran multiple processes per host.
트위터 반응

@xguru: SoundCloud의 아키텍처 변천사 http://t.co/0CmhgHWO 아파치/레일즈/MySQL 에서 시작해 nginx/HAProxy/RabbitMQ/Varnish/멤케쉬디/카산드라/ElasticSearch 등을 붙여간 내역을 잘 정리했네요