SMSRD: A Streaming Graph Data Management System Based on Relational Database

Graphs are one of the most widely used data structures, and with the explosion of data volume, static graph analysis and storage models have become inadequate to meet application demands. This has led to the emergence of dynamic graphs, where graphs exist in the form of streams, leading to the requirements for managing streaming graph data. Existing systems face challenges in simultaneously providing high throughput, data accessibility, and comprehensive streaming data management capabilities, especially under the pressure of handling large data volumes. To address these challenges, we presents a streaming data management system designed upon relational databases. Based on the current application requirements of streaming graph computation and data management, we design the entire framework and unify the top interface for applications, making significant optimizations for high throughput from the aspect of asynchronous and parallel data writing, data storage and sharding strategy. Experimental results validate the system based on relational database outperforms system based on graph database, effectively addressing the demands of streaming graph data management.