Automatic capacity expanding and reducing control method and device, storage medium and processor

Abstract

The invention discloses an automatic capacity expanding and reducing control method. The method includes the steps of periodically monitoring the total number of messages in message queue middleware,the message publishing rate of a message producer and the message consumption rate of each message consumer, triggering a preset distributed application capacity expansion mechanism when the current total message number is greater than a preset maximum message number threshold, and triggering a preset distributed application capacity reduction mechanism when the current total message number is less than a preset minimum message number threshold. According to the method, the capacity expansion and reduction control of the distributed application is automatically realized, the state of higher load and the state of lower load of the message queue middleware are monitored in time, the capacity is reasonably expanded when the load is higher, and the capacity is reasonably reduced when the loadis lower, so that the utilization rate of distributed cluster resources is improved.

Description

technical field [0001] The present invention relates to the technical field of data processing, and more specifically, to a control method, device, storage medium and processor for automatic capacity expansion and contraction. Background technique [0002] Distributed applications will use a message queue as the message middleware architecture. Under this architecture, if the rate of message production is greater than the rate of message consumption, the data will be backlogged in the message queue middleware. The message queue middleware generates load pressure, and the unstable middleware will crash and lose data. [0003] The traditional solution is generally for software developers to manually start one or more copies of distributed applications, thereby increasing the message consumption rate of the message queue middleware, making the message consumption rate greater than the message production rate, so that as few messages as possible are backlogged in the message In...

AI Technical Summary

Problems solved by technology

[0004] However, the problem that often occurs in traditional solutions is that once a backlog of data in the queue is found, it is likely that there is already a large backlog of data, and developers may not be able to have a very accurate judgment on the consumption rate of distributed applications, resulting in more startups The copy of the distributed application leads to the waste of the resources of the entire distributed cluster

Images