Module 5 operation

1.  Computational performance prediction

From the courseware of module 5 ：

【 User volume 】

1. 2020.9 Live every month 5.11 Billion , Diurnal activity 2.24 Billion ( Reference resources 《 Microblogging 2020 User development report 》).

【 Micro-blog 】

Considering that microblog is a business that sees more and sends less , Suppose that the average person sends 1 Micro-blog ( Only consider text microblog ), The daily sending volume of microblog is about 2.5 Billion bars .

Most people tweet in the morning 8:00~9:00 spot , At noon, 12:00~13:00, evening 20:00~22:00, Suppose that the proportion of the total amount of microblogging in these time periods is 60%, Then this 4 The average number of hours of tweeting  TPS  The calculation is as follows :2.5 Billion  * 60% / (4 * 3600) ≈ 10 K/s.

【 Look at micro-blog 】

Because the vast majority of microblog users watch microblogs for big V And stars , Therefore, we assume that the average number of people watching a microblog is 100 Time , The number of times you watch microblog is :

2.5 Billion  * 100 = 250 Billion .

Most people's time period of watching microblog basically coincides with that of sending microblog , So the average number of people watching microblog  QPS  The calculation is as follows :

250 Billion  * 60% / (4*3600) = 1000K/s.

【 Microblog comments 】

Suppose the average number of comments per microblog is 10, The daily number of comments on the microblog is 25 Billion .

• Hot time comments TPS：25 Billion  * 60% / (4 * 3600) ≈ 100 K/s

Suppose the average number of viewers of a comment is 20 Time , The number of times the comment is viewed every day is 500 Billion .

• Hot time comments QPS：500 Billion  * 60% / (4*3600) = 2000K/s

2.  High performance computing architecture for non hot events

The multi-level load balancing architecture of microblog comments can be directly applied to microblogging / Look at the structure of Weibo , Just expand the number of service clusters to the corresponding number ：

• Satisfy  100K TPS：250 platform

• Satisfy  2000K QPS：240 platform

The overall cache architecture is the same as microblogging / Look at micro-blog , I won't repeat .

3.  High availability computing architecture for hot events

Weibo comments on the highly available computing architecture and tweets / Similar to Weibo . The difference is that the comments are not forwarded , But there was a reply . The feature of the reply is that it will not be loaded automatically , So the probability of being read is greatly reduced . But this feature only affects the performance of high-performance computing architecture QPS, It does not affect the highly available computing architecture . Therefore, the highly available computing architecture of the review is the same as the microblog hot events .