You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Jun 16, 2023. It is now read-only.
longdafeng edited this page Oct 14, 2014
·
1 revision
Live 11.11 Room
on the day of 11.11, lots of seller will provide product discount, most of young people will do shopping in www.taobao.com/www.tmall.com, this is a shopping festival. So the trade amount/count is the highest of the years. All guys of Alibaba want to know the KPI as soon as possible, so this is why start the project. All these project are basing on JStorm/Storm. Normally in two minutes, these KPI will be check out.
This page won't give details to how to compute the trade amount/count, PV/UV. just give a simple procedure. Any consumer's operation such as order, pay, cancel order etc. will trigger one message, these message will be sent to dozens of system. Among these system, one is called RocketMq, JStorm will fetch message from RocketMq, then do all kinds of algorithm computation. At last stored the final result to HBase.