Distributed graph processing method, system and storage medium
By using the BSP-GAS coupled computing model in the distributed graph processing system, the problem of large network transmission overhead in iterative computing is solved, and processing efficiency and system performance are improved.
CN113900810BActive Publication Date: 2025-03-18STATE GRID DIGITAL TECHNOLOGY HOLDING CO LTD +1
1 Cites 0 Cited by
Patent Information
- Application Number
- CN202111181778.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-10-11
- Publication Date
- 2025-03-18
- Estimated Expiration
- 2041-10-11
AI Technical Summary
Technical Problem
In a distributed processing architecture based on MapReduce, iterative computing applications generate higher network transmission overhead, resulting in reduced processing efficiency.
Method used
A distributed graph processing method is proposed. By receiving graph processing tasks, it is decomposed into multiple sub-graph processing tasks and distributed to corresponding computing engines. It can perform synchronous iterative graph processing based on the BSP-GAS coupled computing model to reduce network transmission and scheduling overhead.
Benefits of technology
It improves the convergence speed of iterative computing, improves the performance of distributed graph processing systems, and makes it suitable for various types of graph processing tasks.
✦ Generated by Eureka AI based on patent content.
Abstract
The present application proposes a distributed graph processing method, system and storage medium. After receiving a graph processing task for a target business and obtaining graph data associated with the target business, in order to improve processing efficiency, the graph processing task is decomposed into multiple sub-graph processing tasks and distributed to corresponding computing engines. The computing engines respond to the sub-graph processing tasks and perform synchronous iterative graph processing on the graph data corresponding to the sub-graph processing tasks based on the BSP‑GAS coupling computing model. That is, the respective advantages of the BSP computing model and the GAS computing model are used to implement graph computing, which reduces the network transmission and scheduling overhead generated in the iterative computing process, while obtaining a relatively fast convergence speed, thereby improving the performance of the entire distributed graph processing system, making it applicable to various types of graph processing task application scenarios.
Need to check novelty before this filing date? Find Prior Art