Global flow control method and device in cluster environment

Abstract

The embodiment of the invention discloses a global flow control method and device in a cluster environment, relates to the technical field of big data, and can realize a flow control scheme without invading original service codes. The method comprises: after a service request is received, determmning a service method pointed by the service request; detecting whether an annotation associated with the business method exists or not; if the associated annotation exists, detecting whether the value of a counter exceeds a threshold value or not; and if the value of the counter does not exceed the threshold value, executing the service method and adding 1 to the value of the counter, wherein the value of the counter is stored in redis, the redis belongs to a service cluster, the service cluster is used for executing at least two service methods, and when the execution of the service methods is finished, the value of the counter is subtracted by 1. The method and device are suitable for globalflow control of the distributed system.

Description

technical field [0001] The present invention relates to the technical field of big data, in particular to a global flow control method and device in a cluster environment. Background technique [0002] With the rapid development of the e-commerce industry, the business volume of many online usage scenarios has gradually increased, which has promoted the progress of big data technology in disguise. [0003] At present, operators of e-commerce platforms mostly use a single jvm (Java Virtual Machine, virtual machine) flow control solution. It is difficult to further improve the granularity and flexibility of flow control. When encountering many promotional activities, some The large increase in the number of visits to popular products takes up more resources and other products cannot be purchased normally, or there are problems such as excessive pressure on the database due to the sudden increase in the size of the cluster machine and the mismatch with the database carrying cap...

AI Technical Summary

Problems solved by technology

[0003] At present, operators of e-commerce platforms mostly adopt a single jvm (Java Virtual Machine, virtual machine) flow control solution, and the granularity and flexibility of flow control are difficult to further improve. Improvement, when there are many promotional activities, there will be a large increase in the number of visits to some popular products, which will take up more resources and cause other products to be unable to be purchased normally, or the sudden increase in the size of the cluster machine will not match the carrying capacity of the database. Leading to problems such as excessive pressure on the database

[0004] These problems were not prominent a few years ago, but the scale of current promotional activities is getting bigger and bigger, and the business form is becoming more and more complicated. Once these problems appear, It often causes a huge loss of sales every second, so how to further optimize the global flow control scheme has become a research direction in the field of e-commerce

Images