Distributed Storage Data Update Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In distributed storage systems, updating data across replicates is inefficient due to significant network transmission delays caused by large data transmission volumes, especially when multiple updates are batch processed, leading to increased complexity and delay in maintaining data consistency.
Innovation Solution
A method where a sender terminal device receives an update request, separates update data and control information, and sends them separately to a receiver terminal device using data and control information update requests, respectively, to reduce wait times and enhance system performance by processing updates concurrently with data transmission.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple data updates are batch processed in a single update request, then the logic processing efficiency is improved, but the network transmission delay increases significantly
Solution Approach 1:
The patent segments the update request into two separate parts: a control information update request (containing metadata, version information, and update instructions) and a data update request (containing the actual data blocks). This segmentation allows the control information to be transmitted first with minimal overhead, while data transmission can occur concurrently or subsequently without blocking the processing flow, thereby reducing overall network transmission delay while maintaining efficient batch processing capability.
2Reliability
If update requests are processed in serial execution to maintain data consistency, then the consistency is ensured, but the update delay time increases
Solution Approach 1:
The patent implements preliminary action by having the coordinator send the control information update request to slave servers before sending the actual data update requests. The slave servers prepare their environments and acknowledge receipt of control information in advance, allowing the subsequent data transmission to proceed without waiting for sequential processing confirmation, thus reducing update delay while maintaining consistency through pre-established control protocols.
Solution Approach 2:
The patent enables continuous useful action by allowing the coordinator to send multiple data update requests to different slave servers concurrently after the control information is distributed. Instead of waiting for each slave server to complete processing sequentially, the system maintains continuous data transmission and processing across multiple nodes simultaneously, reducing overall update delay while ensuring consistency through the pre-sent control information that coordinates all updates.
Data Source
AI summary
A method for updating data in a distributed storage system is disclosed. The method may comprise receiving, by a sender terminal device, an update request; obtaining, by the sender terminal device, update data and control information from the update request; sending, by the sender terminal device, the update data, via a data update request, to a receiver terminal device; and sending, by the sender terminal device, the control information, via a control information update request, to the receiver terminal device.