Fair Share Scheduler for Shared Storage Resource Scheduling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional shared storage systems lack flexibility and efficiency in allocating resources to applications with varying performance needs, as they prioritize overall throughput over individual application requirements, leading to inefficiencies in I/O operations.
Innovation Solution
A fair share scheduler that balances fairness and I/O efficiency by using adaptive mechanisms to adjust batch sizes and concurrency, allowing for variable batch sizes and bounded concurrency to optimize resource allocation based on workload characteristics, thereby enhancing I/O efficiency while maintaining fairness guarantees.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a shared storage system optimizes overall throughput by treating all requests the same, then overall system throughput is improved, but individual application performance requirements cannot be met
Solution Approach 1:
The patent segments the shared storage system into multiple queues, with each queue dedicated to a specific application or workload type. This segmentation allows the system to maintain overall throughput optimization while providing tailored performance characteristics to individual applications through application-specific scheduling parameters and queue configurations.
Solution Approach 2:
The patent implements dynamic scheduling parameters that can be adjusted based on workload characteristics and system conditions. The scheduler can dynamically modify queue weights, scheduling algorithms, and resource allocation to balance overall throughput with individual application performance needs, making the system adaptable to varying requirements.
2Adaptability or versatility
If dedicated storage resources are allocated to applications desiring higher performance, then individual application performance is improved, but system flexibility and resource utilization efficiency deteriorate
Solution Approach 1:
The patent creates a multi-functional scheduling framework where a single shared storage system can serve multiple applications with different performance requirements. The system universally supports various scheduling algorithms, queue configurations, and performance tiers, allowing dedicated performance allocation without requiring physically dedicated resources for each application.
Solution Approach 2:
The patent utilizes parameter changes in scheduling policies, queue weights, and resource allocation coefficients to provide different performance levels to different applications. By dynamically adjusting these parameters, the system can allocate more resources to performance-critical applications while maintaining efficient overall utilization, without the rigidity of physical resource dedication.
3Productivity
If a scheduler enhances overall throughput by treating all requests equally, then system-wide efficiency is improved, but I/O efficiency for specific applications deteriorates
Solution Approach 1:
The patent applies local quality by implementing application-specific scheduling parameters and queue configurations within the overall scheduling system. Each application queue can have optimized I/O scheduling parameters tailored to its specific workload characteristics, reducing I/O latency for that application while the system maintains overall throughput optimization through coordinated scheduling across all queues.
Data Source
AI summary
To schedule workloads of requesters of a shared storage resource, a scheduler specifies relative fairness for the requesters of the shared storage resource. In response to the workloads of the requesters, the scheduler modifies performance of the scheduler to deviate from the specified relative fairness to improve input/output (I/O) efficiency in processing the workloads at the shared storage resource.