was originally intended to use redis"s HyperLogLog for statistics, but since users only have one chance to like each article, it is impossible to keep playing redis to determine whether users have liked it or not. But the efficiency of using mysql is too low. I don"t know if you have any good idea-sharp-sharp-sharp problem description
be reasonable, why do you say mysql is inefficient?
giving likes is a very, very simple operation. And for the client, the real-time requirement of the data is very low.
these two points determine that the performance and efficiency requirements of the like system are basically equal to no requirements. It can even be said that as long as the system completes the database update of the like behavior within 10 seconds, the vast majority of users will not feel the impact on the experience.
recently, we are going to do a statistics on the activity of the industry in the project, and the three factors that affect the activity of the industry are the number of posts issued by the industry, the number of comments on the information of the indu...
Reading websites often have today s page views, today s posts, today s comments and so on. Which statistical scheme is better? file + Cookie mode, the file on the server is emptied once in 24 hours. the dat file is accessed only once by the client...
city id name 1 Beijing 2 Shanghai 3 Tianjin news Newslist id name city 1 my 1 2 my 2 3 my 2 4 my 3 find out how many messages are there in Beijing how many messages are there in Shanghai how many messages are there in Tianjin ...
there is a sample, column an is the number, column b is the number of occurrence of the number, the sample is very large, the standard deviation and normal distribution map of this sample is required, how should we deal with it with python? I ve checked...
I write a lot of api, myself. These api may be called by many three parties. How should I count the situation of each third party calling api? should the statistics be updated before each call to the API, and then return the interface data? or should mul...