Network Data Caching via Predictive Collection to Reduce Latency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Centralized management of remote computers often experiences latency due to infrequent data access and inefficient network bandwidth usage, as existing methods either rely on caching only requested data or continuous polling of remote systems.
Innovation Solution
Collecting and caching additional data beyond the immediate request from remote computers, reducing the need for frequent remote connections by using a data manager that dynamically reallocates resources and requests configuration and utilization data from providers, including workload managers and user interfaces, which conforms to protocols like WBEM or SNMP.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If data is cached only upon request from remote computers, then network bandwidth is conserved, but latency increases when data is needed
Solution Approach 1:
The patent applies preliminary action by collecting and caching related data from remote computers in advance, before it is actually requested. The data collector proactively gathers data based on prediction of future requests, storing it in the local database. This eliminates the need to wait for requests before caching data, thereby reducing latency while avoiding continuous polling since data is collected only when prediction indicates future need.
2Loss of information
If continuous polling is used to keep local data up-to-date, then data availability is improved, but network bandwidth is consumed by superfluous accesses
Solution Approach 1:
The system applies self-service by using the data collector to autonomously determine when and what data to collect based on prediction algorithms. Instead of relying on continuous external polling or reactive collection only after requests, the system services itself by predicting its own data needs and proactively collecting relevant data, thereby maintaining data availability without superfluous network accesses.
Solution Approach 2:
The patent changes the parameter of data collection frequency from continuous (polling) or reactive (on-request) to predictive (conditional). The system dynamically adjusts when to collect data based on prediction outcomes, transforming the collection pattern into an optimized sequence that balances data freshness with network efficiency.
3Loss of energy
If only requested data is collected from remote computers, then network bandwidth is optimized, but the number of remote connections must increase to meet multiple data requests
Solution Approach 1:
By collecting related data in advance based on prediction, the system performs preliminary actions that consolidate multiple potential requests into single collection operations. This reduces the total number of remote connections needed, as related data is gathered proactively before requests arrive, improving connection efficiency without sacrificing bandwidth optimization.
Data Source
AI summary
The present invention provides a system and method for managing a network. In response to an original request, a management station sends an augmented request to a remote server for information fulfilling the original request plus additional information available.