High performance and high availability computing architecture of "microblog comments" in microblog system

【 High performance computing architecture for non hot events 】

【 User volume 】

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

【 User behavior modeling 】

【 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.

【 Comment on Weibo 】

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%, Comments depend on tweeting and reading .

According to the estimate, I read the microblog QPS, Suppose the people who read Weibo have 25% Probability comment Weibo , Then calculate the average of comment microblogs  TPS The calculation is as follows ：

2.5 Billion  * 60% / (4 * 3600) * 25% ≈ 2.5 K/s.

【 Business characteristic analysis 】

Comment microblog is a typical writing operation , So you can't use cache , Load balancing can be used .

【 Architecture analysis 】

Over 100 million users , You should use a multi-level load balancing architecture , Cover  DNS -> F5 -> Nginx ->  Multilevel load balancing of gateway .

【 Architecture design 】

• The load balancing algorithm depends on the login status when it chooses to comment on Weibo , The login status is generally saved in the distributed cache , So when commenting on Weibo , Send the request to any server , Choose here “ polling ” perhaps “ Random ” Algorithm .

• Estimating the number of business servers commenting on microblogging involves several key processes ： Content review （ Rely on the audit system ）、 Write data to storage （ Dependent on storage system ）、 Data write cache （ Rely on the cache system ）, So, per service per second 500 To estimate , complete 2.5K/s Of  TPS, need 5 Servers , Plus a certain amount of reserve ,8 This server is almost .

【 Multilevel load balancing architecture 】

【 High availability computing architecture for hot events 】

【 User behavior modeling 】

It's hard to predict , Related to the influence and scope of the event .

【 Business characteristic analysis 】

• Comment on Weibo
• Business logic is basically equivalent to microblogging , Its importance and influence are not as important as the original Weibo .

【 Architecture design analysis 】

Core architecture design idea ： Since it is impossible to predict , Then take precautions ！

• Comment on Weibo comment is less important and influential than the original Weibo , Think about it “ Comment on Weibo ” Current limiting , But the request cannot be discarded , Consider using “ Leaky bucket algorithm ”.

【 Sketch Map 】