Virtual Storage Manager Performance Leveling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing storage systems fail to optimally allocate non-volatile memory resources among multiple hosts with varying workloads in multi-tenant systems, leading to suboptimal performance due to differing interface connections and network load variations.
Innovation Solution
A virtual storage manager dynamically adjusts the operational profiles of data storage devices based on their processing and interface speeds, changing operating parameters to maintain target aggregate performance by selecting between high performance and maintenance profiles depending on network load thresholds.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If storage devices are connected through networks with variable load, then system flexibility and scalability are improved, but interface speed and latency become unpredictable
Solution Approach 1:
The patent implements dynamic operational profiles that adjust storage device performance parameters in real-time based on network load conditions. The system transitions between different operational states (e.g., performance-optimized, latency-optimized, power-saving) to adapt to changing interface speeds, thereby maintaining system flexibility while compensating for speed variability.
Solution Approach 2:
The system changes operational parameters such as cache allocation, I/O scheduling policies, and background process priorities based on detected interface speed conditions. When network load increases and interface speed decreases, the system adjusts parameters to optimize for the current state, resolving the contradiction between adaptability and speed consistency.
2Productivity
If storage resources are dynamically allocated among multiple hosts, then resource utilization efficiency is improved, but performance consistency across different hosts deteriorates
Solution Approach 1:
The patent applies different operational profiles to storage devices based on their specific interface characteristics and host requirements. Each host-reachable storage device can have customized performance parameters adjusted according to its local conditions (interface type, network load, host workload), enabling efficient resource allocation while maintaining performance consistency through localized optimization.
Solution Approach 2:
The system continuously monitors interface speeds, network load conditions, and host performance requirements, using this feedback to dynamically adjust operational profiles. This closed-loop control ensures that resource allocation efficiency is maintained while performance consistency is preserved through real-time adaptations based on actual system state.
3Adaptability or versatility
If background processes are executed on storage devices, then data management functionality is improved, but available processing speed for host operations decreases
Solution Approach 1:
The system implements periodic background processes that execute data management tasks (garbage collection, wear leveling, metadata updates) during intervals when host I/O demand is lower. By scheduling these operations periodically rather than continuously, the system maintains comprehensive data management functionality while minimizing the impact on host processing speed during active periods.
Solution Approach 2:
The operational profiles dynamically control the allocation of processing resources between background tasks and host operations. When interface speed indicates high network load or host demand, the system reduces background process intensity or suspends non-critical tasks, thereby maintaining data management functionality at necessary levels while preserving processing speed for host operations.
Data Source
AI summary
Systems, methods, and data storage devices for using data storage device operational profiles for interface-based performance leveling are described. Data storage devices are connected to a virtual storage manager using network or storage bus connections. Comparisons of the data processing speed and interface speed of each data storage device and its connection may be used to determine an active device operational profile for each data storage device to reach a target aggregate performance. The device operational profiles may be sent to the data storage devices to change their operations, such as the handling of background processes.