RMPF: Real-Time Message Processing Framework on Multi-Service Domains

Abstract

In the world of real-time computing, there is a demand for platforms that deliver both speed and efficiency while conserving hardware resources. High reliability applications typically use dedicated platforms for their inherent reliability, but they are more expensive. On the other hand, traditional general-purpose platforms are cheaper but suffer from wasted hardware resources when processing messages, hindering distributed processing and process monitoring. To address these limitations, this paper presents the Real-time Message Processing Framework on Multi-service Domains (RMPF), an innovative integrated platform for real-time distributed message processing on multi-service domains. The platform supports a scalable and flexible multi-layer architecture by providing modularity that allows each layer to be designed and updated independently. The interaction between each layer enables real-time distributed message processing and distributed management of multi-service processing in a city-main based, high-speed distributed environment. By accelerating message processing, using a dedicated buffer pool created for each application process, and monitoring the status of all messages in the buffer, the high reliability inherent in dedicated platforms is ensured. In addition, it minimizes the waste of hardware resources by running application processes as daemons in buffer pools. In particular, to provide domain-based multiservice, it supports RMPF grouping by task to distribute tasks to appropriate domains to eliminate system bottlenecks. It is also implemented in a modular way to achieve both scalability and adaptability through various multiservice distributed management schemes such as multiservice API, recovery, monitoring, and multiservice provisioning.

Keywords

• Multi service
• distributed message processing
• unified platform
• memory management
• multi domain