Socket Allocation for File Transfer Time Equalization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Distributed computer systems face challenges in ensuring consistent file transmission across multiple web servers due to reliance on a single repository, which results in low reliability and scalability issues caused by varying network latency.
Innovation Solution
A system and method that determine the number of sockets for file transfer based on measured file transfer rates to web servers, using more sockets for slower connections and fewer for faster ones to equalize transfer time, ensuring consistent file availability across all servers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If a single shared repository is used for file storage, then file consistency across web servers is improved, but system reliability and scalability deteriorate due to single point of failure and limited bandwidth
Solution Approach 1:
The patent divides the single shared repository into multiple distributed storage nodes across different web servers. Each server maintains local file copies, eliminating the single point of failure while preserving file consistency through coordinated updates using two-phase commit protocol and checksum verification.
Solution Approach 2:
The patent introduces a file distribution service as an intermediary that manages file updates across the distributed system. This service coordinates the two-phase commit process, generates MD5 checksums for verification, and ensures that all storage nodes receive consistent file updates without relying on a single central repository.
2Manufacturing precision
If a single shared repository is used for file storage, then file consistency across web servers is improved, but system scalability deteriorates due to limited bandwidth and finite response time
Solution Approach 1:
The patent segments the centralized repository into multiple distributed storage locations across different web servers. This segmentation increases system bandwidth capacity and improves response time by allowing parallel file operations, while file consistency is maintained through the two-phase commit protocol and checksum-based verification mechanisms.
Solution Approach 2:
The patent transitions from a single-dimensional centralized storage model to a multi-dimensional distributed storage architecture. Files are replicated across multiple servers and storage nodes, creating additional dimensions for data access and improving system scalability while maintaining consistency through coordinated update protocols.
3Manufacturing precision
If two-phase commit procedure is used to update files across storages, then file consistency is improved, but transfer time increases due to waiting for all servers to be ready
Solution Approach 1:
The patent implements a preliminary preparation phase in the two-phase commit protocol where the file distribution service first validates the file (generating MD5 checksums) and prepares update instructions before actual file transfer. This preliminary action ensures that when the commit phase begins, all servers are pre-configured and ready, reducing the actual transfer time while maintaining consistency.
Solution Approach 2:
The patent replaces the traditional synchronous two-phase commit mechanism with an optimized approach that uses asynchronous file transfer combined with MD5 checksum verification. This substitution allows files to be transferred in parallel to multiple servers simultaneously, rather than sequentially waiting for each server's readiness, significantly reducing total transfer time while maintaining consistency through cryptographic verification.
4Loss of time
If multiple sockets are used for file transfer to equalize transfer time, then transfer time consistency is improved, but system complexity increases
Solution Approach 1:
The patent implements a feedback mechanism where the file distribution service monitors network conditions and server response times, then dynamically adjusts the number of sockets used for file transfer to each server. This feedback loop equalizes transfer times across servers while automatically managing socket allocation, reducing the perceived complexity through adaptive control rather than manual configuration.
Solution Approach 2:
The patent dynamically changes the number of sockets parameter based on measured network conditions and server performance characteristics. By adjusting this parameter adaptively, the system achieves equalized transfer times across multiple servers without requiring complex manual socket management, as the file distribution service automatically optimizes socket allocation based on real-time feedback.
Data Source
AI summary
System, method and program for transferring same files from a first computer to a plurality of second computers. Respective measures of file transfer rates from the first computer to the second computers are determined. Based on the respective measures of file transfer rates from the first computer to the second computers, determinations are made as to respective numbers of sockets within the first computer to transfer the files to the second computers to tend to equalize the time required to transfer the files from the first computer to each of the second computers. The files are transferred to each of the second computers using the determined number of sockets, wherein for each of the second computers where the number of sockets is greater than one, the files are divided between the number of sockets for transfer to the respective second computer. A greater number of sockets are used for those of the second computers for which the first computer has slower file transfer rates than for those of the second computers for which the first computer has greater file transfer rates.