Structure the fifth operation of the actual camp module

Comment on Weibo computing performance estimation

1.  Estimated number of users

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

2.  Key behaviors

  Write a comment 、 Respond to others' comments 、 Delete your comments

  See the comments

3.  Performance prediction

 （1） Write a comment 、 Respond to others' comments 、 Delete your comments

  Because most Weibo users read / The object of commenting on Weibo is big V And stars , So let's assume that there are 10% Will comment , So the average number of comments on a microblog is 10 Time , The number of comments on microblog is ：2.5 Billion  * 10 = 25 Billion .

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%, Most people comment on Weibo for the same time period as they post , So the average number of comments on Weibo  TPS  The calculation is as follows ：25 Billion  * 60% / (4*3600) = 100K/s.

 （2） See the comments

  Because the vast majority of microblog users read microblog comments on big V And stars , Therefore, we assume that the average number of people watching a microblog is 100 Time , Comments are requested together with Weibo , The number of times to watch microblog comments is ：2.5 Billion  * 100 = 250 Billion .

  Most people read microblog comments in the same time period as they send microblogs , So the average number of microblog comments  QPS  The calculation is as follows ：

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

4.  Whether to split independent services

  There is a big difference between writing comments and reading comments , For example, you can't use cache to write comments , Reading comments depends on caching . It is divided into write comment service and read comment service .

High performance computing architecture for non hot events

1.  Write a comment 、 Respond to others' comments 、 Delete your comments

 （1） Business characteristic analysis

  All write operations , I hope I can see it in time , So you can't use cache , Load balancing can be used .

 （2） Architecture analysis

  The number of requests reached onebillion , One hundred thousand seconds , Use a multi-level load balancing architecture , Cover  DNS -> F5 -> Nginx ->  Multilevel load balancing of gateway .

 （3） Architecture design

  Load balancing algorithm selection

  Depending on login status when posting comments , The login status is generally saved in the distributed cache , So when commenting , Send the request to any server , Choose here “ polling ” perhaps “ Random ” Algorithm .

  Estimate the number of business servers

  Commenting 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 100K/s Of  TPS, need 200 Servers , Plus a certain amount of reserve ,250 This server is almost .

2.  See the comments

 （1） Business characteristic analysis

  Reading comments is a typical reading scene , Because the comments can't be modified after they are sent , Therefore, it is very suitable to use cache architecture , At the same time, due to the large number of requests , Load balancing architecture also needs .

 （2） Architecture analysis

  The number of requests reaches 10 billion , Seconds concurrent millions , Use a multi-level load balancing architecture , Cover  DNS -> F5 -> Nginx ->  Multilevel load balancing of gateway . You should use a multi-level cache architecture , In especial  CDN  cache , Is the core of cache design .

 （3） Architecture design

  Load balancing algorithm selection

  Visitors can directly read microblog comments , Therefore, sending a request to any server can , Choose here “ polling ” perhaps “ Random ” Algorithm .

  Estimate the number of business servers

  hypothesis  CDN  Capable of carrying 90% Of users , So the rest 10% The request to read microblog comments enters the system , Ask for  QPS  by 1000K/s * 10% = 100K/s, Because the processing logic of reading microblog comments is relatively simple , Mainly read cache system , Therefore, it is assumed that the processing capacity of a single business server is 1000/s, Then the number of machines is 100 platform , according to 20% Reserved quantity of , The final number of machines is 120 platform .

3.  Microblog high performance computing solution  -  Overall architecture design

Task assignment  -  Double room 、 Three machine rooms

Task breakdown  -  Write a comment 、 See the comments

Split different services

The overall architecture of multi-level load balancing of microblog comments

See the overall multi-level cache architecture of microblog comments

High availability computing architecture for hot events

1.  Write a comment

  Hot events happen , The number of comments will increase significantly . Because in the business of posting comments , The write buffer is introduced , There is no danger of system downtime , But the user experience will decline , Because it will take longer for you to get your comments .

2.  See the comments

  Hotspot Events can cause cache hotspot problems , Consider multi copy caching , Because the original cache architecture has adopted “ Cache in application , In general , Cache hot issues are not necessarily prominent .