Data Storage Resource Allocation for Data Flow Contention
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data storage resource allocation methods in data flow systems lead to inefficiencies, such as high memory requirements and increased CPU interrupt costs due to frequent allocation and deallocation requests, resulting in decreased processing speed, especially when handling large data volumes.
Innovation Solution
A resource allocation apparatus and method that acquires attributes indicating data storage requirements for data flows, generates a data flow relationship graph to identify potential access contention, and determines resource allocation to prevent contention, thereby optimizing the allocation of data storage resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If data storage resources are allocated to each data flow individually (as in Cantata), then processing speed is improved due to dedicated resources, but memory usage becomes enormous
Solution Approach 1:
The patent applies universality by allowing a single data storage resource to serve multiple data flows simultaneously. The resource allocation unit dynamically assigns the same storage resource to different data flows based on their current access patterns and requirements, enabling one resource to perform multiple functions and serve multiple purposes, thereby reducing total memory usage while maintaining processing efficiency.
Solution Approach 2:
The patent implements dynamics through dynamic resource allocation. Instead of static one-to-one mapping, the system continuously monitors data flow characteristics and dynamically adjusts which data flows share which storage resources. This dynamic allocation allows the system to optimize memory usage in real-time while ensuring that processing speed requirements are met when needed.
2Quantity of substance
If data storage resources are dynamically acquired and collected (as in DirectShow), then memory usage is reduced through resource reuse, but processing speed decreases due to frequent allocation requests interrupting the CPU pipeline
Solution Approach 1:
The patent applies preliminary action by pre-establishing resource sharing relationships and allocation rules before data processing begins. The system analyzes data flow characteristics in advance and configures which resources will be shared and under what conditions, thereby avoiding the need for frequent runtime allocation requests that interrupt the CPU pipeline, while still achieving memory efficiency through reuse.
3Reliability
If many data storage resources are allocated to multiple data flows, then resource reuse is reduced, but access contention is minimized
Solution Approach 1:
The patent implements feedback through continuous monitoring of data flow access patterns and resource usage metrics. The resource allocation unit receives feedback about current system state, including which data flows are actively accessing resources and at what rates, and uses this feedback to dynamically adjust resource allocation decisions, optimizing the balance between preventing access contention and minimizing total resource size.
Data Source
AI summary
There is provided with a resource allocation apparatus. An attribute indicating a requirement for a data storage resource to be allocated to a plurality of data flows to which the attribute is provided beforehand is acquired. A data flow relationship graph indicating a relationship between the plurality of data flows which potentially lead to access contention in the data storage resource is generated. Based on the attribute and the data flow relationship graph, allocation of the data storage resource to the plurality of data flows is determined such that no access contention occurs.