Storage Tier Reallocation Based on Host Performance Bottlenecks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing hierarchical storage technologies fail to optimize storage tier allocation based on host computer performance bottlenecks, leading to inefficient use of high-performance tiers and suboptimal response times, as they do not consider CPU or network utilization rates when reallocating data.
Innovation Solution
A computer system management apparatus that manages multiple host computers and storage tiers by using performance information to determine load thresholds, create reallocation plans, and instruct storage reallocation, ensuring that high-performance tiers are allocated to hosts with performance bottlenecks, thereby optimizing response times across the system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If high-performance storage tiers are allocated to hosts with high I/O loads, then I/O response time is improved, but overall system response time cannot be optimized because non-I/O processing becomes the bottleneck
Solution Approach 1:
The invention changes the allocation parameters from solely I/O-based metrics to multi-dimensional performance metrics including CPU utilization rate, network utilization rate, and I/O distribution. This allows the system to identify hosts where non-I/O processing is the bottleneck and prevent allocating high-performance storage tiers to them, thereby optimizing overall system response time rather than just I/O response time
Solution Approach 2:
The management apparatus continuously monitors host computer performance information and uses this feedback to dynamically adjust storage tier allocations. By measuring current system state and comparing it against allocation criteria, the system can reallocate storage resources to match actual performance bottlenecks, ensuring that high-performance tiers are allocated only when they will actually improve overall system response time
2Speed
If high-performance storage tiers are allocated to all hosts, then I/O processing speed is maximized, but storage resource utilization efficiency deteriorates
Solution Approach 1:
The invention applies different storage tier qualities to different hosts based on their specific performance characteristics. Instead of uniformly allocating high-performance tiers to all hosts, the system analyzes each host's CPU utilization, network utilization, and I/O distribution to determine which hosts would actually benefit from high-performance storage. This localized allocation strategy ensures that high-performance resources are concentrated where they are most needed, maximizing both I/O processing speed and resource utilization efficiency
Data Source
AI summary
An object of the present invention is to efficiently use a hierarchical pool. A management server manages performance information of multiple host computers, and based on each performance information, determines whether a prescribed host computer, which comprises a load of equal to or larger than a preconfigured prescribed threshold, exists among the multiple host computers. The management server creates a reallocation plan, which stipulates an allocation amount of a real storage area for each of multiple storage tiers, with respect to a prescribed virtual logical volume used by the prescribed host computer. Based on the reallocation plan, the management server decides a corresponding relationship between each logical storage area and each real storage area of each storage tier, and notifies a storage apparatus of this corresponding relationship.