Generally, Java developers like to use synchronous code to write programs, because this request/response method is relatively simple and more in line with programmers’ thinking habits; this approach is good until the system has a performance bottleneck. When using synchronous programming, since each thread can only initiate one request at the same time and wait for the return synchronously, in order to improve system performance, we need to introduce more threads to achieve parallel processing. However, when accessing shared resources under multiple threads, resource contention and concurrency problems will inevitably be introduced; in addition, the number of threads is limited at the operating system level, and it is impossible to increase the number of threads infinitely to provide system performance; and , The use of synchronous blocking programming will also waste resources. For example, when a network IO request is initiated, the calling thread will be in a state of synchronous blocking waiting for the response result, and at this time the calling thread can obviously do other things, and wait for the network IO response result to return Process the results later.

It can be seen that parallel programming by increasing the number of threads in a stand-alone system is not a "magic bullet". By writing asynchronous, non-blocking code, you can use the same underlying resources to switch execution to another active task, and then return to the current thread to continue processing after the asynchronous processing is completed, thereby improving system performance.

Asynchronous programming is a method that allows programs to run in parallel. It allows a unit of work in the program to run independently of the main application thread, and after the work unit is finished, the main application thread will be notified of its running result or Reason for failure. Using asynchronous programming can improve application performance and responsiveness.

For example, when the calling thread initiates a network IO request in an asynchronous manner, the calling thread will not synchronously block and wait for the response result, but after the request context is saved in the memory, it will immediately return to do other things, and then use the IO after the network IO response result is returned. The thread notifies the business thread that the response result has been returned, and the business thread processes the result. It can be seen that the asynchronous call method improves the utilization of threads, allowing the system to have more thread resources to handle more requests. For example, in a mobile application, after the user operates the mobile device screen to initiate a request, if it is synchronously waiting for the background server to return the result, when the background service operation is very time-consuming, it will cause the user to see the mobile device screen freezing (always in the request processing In), the user cannot operate other functions of the mobile device before the result is returned, which is very bad for the user experience. When using asynchronous programming, when a request is initiated, the calling thread will return immediately, and the specific return result will be rendered asynchronously through the UI thread, and the user can use other functions of the mobile device during this period.

This study note shared today is a top tutorial note on Java asynchronous programming!

Due to the excessive content of the notes, only part of the content can be shown below for everyone. The free way to obtain the full Blu-ray version is at the end of the article!

Detailed chapter list