System and method for capacity evaluation of event analysis engine adopting complex event processing cloud platform

Abstract

The invention relates to a system and a method for capacity evaluation of an event analysis engine adopting a complex event processing cloud platform. The system comprises an event analysis engine, an event source management module and an analysis engine capacity evaluation module, wherein the event analysis engine is used for receiving an external event and loading various event processing rules; the event source management module is used for managing event metadata in the complex event processing cloud platform and providing an event structure corresponding to each event type; the analysis engine capacity evaluation module is used for carrying out capacity evaluation of the event analysis engine according to the arrived event quantity and the event types in the event analysis engine, and the event structure corresponding to each event type in the event source management module. By adopting the system and the method for capacity evaluation of the event analysis engine adopting the complex event processing cloud platform in the structure, capacity evaluation is carried out on the basis of a single node of event throughput, namely a single java virtual machine (JVM) process, so that the overall operational level of the complex event processing platform is greatly improved, and the method is stable and reliable in working performance, and wide in application range.

Description

technical field [0001] The present invention relates to the field of computer software, in particular to the field of enterprise-level memory computing and high real-time event flow analysis system, and specifically refers to a system and method for evaluating the capacity of an event analysis engine on a complex event processing cloud platform. Background technique [0002] With the advent of the big data era, the storage and analysis of massive data has gradually become one of the main development directions of information technology development. For massive data processing, it can be divided into non-real-time processing technology before or after the event, and real-time processing technology during the event. Among them, the real-time processing technology for in-process processing requires high computing time, so it needs to be completed by pure memory computing. As for pure memory computing, with the decline of memory prices, large-scale distributed memory computing ...

AI Technical Summary

Problems solved by technology

On the one hand, in the memory heap management of the JVM, objects are divided into young generation, old generation, and permanent generation. Events in complex event processing scenarios and the Class class of the program itself will become managed objects in the JVM memory heap ; On the other hand, the JVM's automatic memory recovery mechanism causes the memory space to not be released immediately after the event processing is completed. The fundamental reason is that it is impossible to distinguish the memory space occupied by the program itself from the space occupied by the event object

Based on the above two reasons, the JVM process can only simply provide the ability to query the memory heap occupancy, so it is impossible to distinguish real events from the space occupied by the Java program's own Class class through the query results, and it is also impossible to know the real-time memory usage in real time. Due to the situation, it is difficult to calculate the remaining event capacity of the event analysis engine in this scenario, and it is impossible to evaluate the capacity of single-process nodes

[0008] At the same time, in the cloud platform environment, the overall capacity of the platform is composed of a single node, and the lack of single-node capacity evaluation capabilities will lead to the lack of overall platform capacity evaluation capabilities

Images