Interface middleware system

Abstract

The invention provides an interface middleware system, which relates to the technical field of cloud computing database architectures, and comprises: a model management module for storing a pluralityof data models and data markets; a first interface management module which is used for receiving a service connection request sent by each service system through a first universal standard interface and decomposing each service connection request into a plurality of Hadoop processing transactions; a transaction processing module which comprises a judgment unit used for generating a corresponding first execution instruction when each Hadoop processing transaction represents that a corresponding service connection request is a periodic statistical query, and generating a corresponding second execution instruction when each Hadoop processing transaction represents that the corresponding service connection request is not the periodic statistical query, and the second interface management module used for sending the first execution instruction and the second execution instruction to the Hadoop cluster server through a second standard interface. According to the method, the difficulty and period of developing the distributed big data application by the upper-layer application are effectively reduced, and the migration and development cost is reduced.

Description

technical field [0001] The invention relates to the technical field of cloud computing database architecture, in particular to an interface middleware system. Background technique [0002] At present, the telecommunications industry generally faces the problem of difficult management and analysis of massive business data. After years of research and discussion in the telecommunications industry, two main lines of development have basically been formed on how to effectively utilize the massive data of the huge telecommunications business. The first is to build a cloud service platform to provide cloud computing services to customers. The second is to build a massive data support platform to provide big data analysis services for telecom's internal business systems. [0003] As the concept of traffic management has gained consensus in the industry, analysis and mining based on user behavior has become a very important direction for the telecom industry to improve its operati...

AI Technical Summary

Problems solved by technology

[0004]Hadoop platform is an open source software with many components. It is currently the most widely used big data platform. However, the application cycle based on Hadoop platform is long, costly and technically difficult , which requires high ability and experience of application developers, so its powerful functions also bring a lot of development costs and learning costs

At present, many internal systems of operators are developed based on traditional relational databases, and the cost of migration is relatively high.

Moreover, as a unified data storage and computing resource, the Hadoop platform needs to be open to multiple upper-layer applications, so there may be resource competition between multiple applications accessing the same data resources.

Images