It is said that mongodb’s MapReduce is single-threaded and has poor performance. What is going on? How bad is it? ? Who can tell us the principle?
I don’t know if the execution is single-threaded, but if it is a production environment, it is better not to directly access the mapReduce results each time. Depending on the amount of data, it will still take some time. Our data is ten million levels, and it takes about 5-6 seconds to execute mapReduce every time. Fortunately, our application is not very real-time. So basically it is to cache the data for 2 hours and then execute mapReduce to get the latest results.