Policy Manager Optimizes Cloud Data Transmission
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The performance of cloud-based virtual computing systems is hindered by slow data transmission between local clients and remote applications, particularly due to low bandwidth and lossy wireless connections, leading to sluggish user experiences.
Innovation Solution
Implementing a virtual network tunnel with forward error correction (FEC) and overlay management, managed by a policy manager that configures communication paths based on bandwidth and error characteristics, to optimize data transmission and reduce delays.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data transmission is performed over a slow or lossy wireless connection between local client and cloud-based application, then cloud-based virtual computing can be accessed, but user perceived performance becomes sluggish
Solution Approach 1:
The system performs preliminary actions by pre-fetching data blocks from the cloud-based file system before they are actually needed for application execution. The policy manager predicts which data blocks will be needed and transfers them in advance over the network connection, reducing the perceived latency when the application needs to access the data.
Solution Approach 2:
The patent introduces an intermediary component (the policy manager and optimization module) between the local client and the cloud-based application. This intermediary manages data transmission by buffering, pre-fetching, and optimizing network I/O operations, thereby masking the slow network performance from the application and user.
2Speed
If data is cached locally to improve access speed, then data retrieval is faster, but memory resources on the local client are consumed
Solution Approach 1:
The system applies local quality by differentiating between hot data (frequently accessed) and cold data (rarely accessed). The policy manager maintains optimized local caches for hot data blocks that are predicted to be needed soon, while allowing cold data to remain in the cloud-based file system. This selective caching approach maximizes performance benefits while minimizing memory resource consumption.
Solution Approach 2:
The caching strategy is dynamic rather than static. The policy manager continuously monitors application behavior patterns and adjusts which data blocks are cached locally based on predicted future access needs. Data blocks are dynamically added to or removed from the local cache based on changing application requirements and available memory resources.
3Productivity
If multiple data connections are established to improve throughput, then bandwidth utilization increases, but connection management complexity increases
Solution Approach 1:
The policy manager implements self-service by automatically managing multiple data connections without requiring complex external orchestration. The system autonomously establishes, maintains, and optimizes multiple parallel network connections to the cloud-based file system, dynamically allocating data transfer tasks across connections based on their current performance characteristics.
Solution Approach 2:
The patent segments data transfer operations across multiple independent connections. Instead of using a single connection for all data I/O, the system divides data blocks and distributes their transfer across multiple parallel connections, thereby increasing overall throughput while isolating connection management complexity into manageable segments handled by the policy manager.
Data Source
AI summary
A method includes establishing a communication link between a communication interface of a first computing device and a virtual machine executed by a second computing device, receiving an available bandwidth parameter associated with the communication link, receiving link characteristic data associated with the communication interface, and configuring a characteristic of the communication link based on the available bandwidth parameter and the link characteristic data.A system includes a first computing device including a communication interface and executing a policy manager and a second computing device executing a virtual machine. The policy manager is to receive an available bandwidth parameter associated with the communication link, receive link characteristic data associated with the communication interface, and configure a characteristic of the communication link based on the available bandwidth parameter and the link characteristic data.