Timeout processing method and device for client request

Abstract

This application discloses a method and device for overtime processing of client requests. The method includes: the server encapsulates each connection request currently received into a task, and adds it to the pending request task pool and corresponds to the delay threshold of the task. In the timeout monitoring queue of the server, the timeout monitoring queue in the server corresponds to the supported delay threshold one by one, and the tasks in each timeout monitoring queue are arranged in ascending order according to the timeout time; when the pending request task pool is not empty, take out the task from it , trigger the execution process of the task according to the task status; for each timeout monitoring queue, according to the preset timeout scanning cycle, find out the timeout task in the queue and delete it, and put the timeout task into Timeout processing thread pool; when the timeout processing thread pool is not empty, take out the timeout task from it, and trigger the timeout processing process of the task according to the task status. By adopting the invention, the overtime request can be processed in time.

Description

technical field [0001] The invention relates to computer application technology, in particular to a method and device for overtime processing of client requests. Background technique [0002] With the rapid development of the Internet, the number of WEB application systems continues to increase, and the subsequent emergence of distributed systems leads to the intersection of services between various systems, which will involve the mutual calling of interfaces between systems. Especially systems involving different companies, such as between shopping system and warehouse system, shopping system and payment system, etc. The interaction between the external system and the internal system generally adopts the HTTP interface. With the increasing number of users, the number of interactions between the interfaces is also increasing. Correspondingly, the access pressure of the system is also increased. In this case, when the connection request arrives at the server, there will be a...

AI Technical Summary

Problems solved by technology

[0004] In the above timeout processing scheme, since the timeout time of different requests in the queue may be different and has nothing to do with the order of the requests in the queue, that is to say, any request in the queue will reach the timeout time, so , the timing thread needs to traverse the entire request queue. When the length of the queue reaches tens of thousands, it will take a long time to traverse it. In addition, the judgment and processing time of each data will lead to a very time-consuming judgment of a queue timeout.

In addition, the length of each request queue is limited. If the timeout request is not processed in time and the timeout request cannot be deleted from the queue in time, it will appear: due to the existence of the timeout request, the queue is full and new requests cannot be added to The situation in the queue, in this way, will cause subsequent requests to be rejected, resulting in a very poor user experience

[0005] It can be seen that the existing timeout processing scheme has the problem that the timed out request in the request queue cannot be processed in time, which affects the acceptance of subsequent requests, and then affects the service performance of the system

Images