Cluster Parity Assignment Based on Write Speed Ratios
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current fault tolerance implementations in clustered storage systems do not consider varying characteristics of systems, leading to suboptimal write performance and storage space utilization.
Innovation Solution
Determine the write speed for each system in a cluster, calculate the ratio of write speeds, and update parity assignments based on these ratios to achieve a write-balanced distribution of parity data across systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If equal parity assignments are used across all systems in the cluster, then fault tolerance is maintained, but write performance deteriorates due to imbalance when systems have different write speeds
Solution Approach 1:
The patent applies local quality by assigning different parity weights to different systems based on their individual write speeds. Each system receives a customized parity assignment proportional to its write capability, rather than applying a uniform parity assignment across all systems. This resolves the contradiction by tailoring the fault tolerance mechanism to match local system characteristics, thereby maintaining reliability while optimizing overall write performance.
Solution Approach 2:
The patent changes the parameter of parity assignment from a fixed equal distribution to a dynamic distribution based on write speed ratios. By calculating the ratio of write speeds among systems and adjusting parity assignments accordingly, the system adapts to varying performance characteristics. This parameter change enables the system to maintain fault tolerance while avoiding write performance degradation caused by forcing equal parity assignments on systems with different capabilities.
2Productivity
If parity assignments are optimized for write balance, then write performance improves, but storage space utilization deteriorates due to uneven distribution
Solution Approach 1:
The patent implements feedback by continuously monitoring write speeds of systems in the cluster and using this information to adjust parity assignments. The system calculates write speed ratios and dynamically rebalances parity distribution to maintain optimal storage space utilization. This feedback mechanism resolves the contradiction by ensuring that parity assignments are continuously optimized based on actual system performance, preventing both write performance degradation and storage space waste.
3Device complexity
If static parity assignments are used, then system complexity is reduced, but adaptability deteriorates when system characteristics change over time
Solution Approach 1:
The patent applies dynamics by transitioning from static parity assignments to dynamic parity assignments that automatically adjust to changing system conditions. The system periodically recalculates write speed ratios and rebalances parity distributions to match current system capabilities. This dynamic approach resolves the contradiction by enabling the system to adapt to changing write rates and system characteristics while maintaining manageable complexity through automated adjustment mechanisms.
Data Source
AI summary
According to one embodiment, a computer-implemented method includes determining a write speed for each of a plurality of systems in a cluster, determining a ratio of the write speeds for each of the plurality of systems in the cluster, and updating parity assignments to each of the plurality of systems in the cluster, based on the ratio of the write speeds.