Storage System Performance Bottleneck Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current storage system management technologies fail to promptly and accurately identify problematic resources affecting response performance, leading to inefficiencies in maintaining storage subsystem responsiveness to host computers.
Innovation Solution
The storage system employs a performance management method that limits the presentation of resources to administrators by considering I/O request status and mapping information, enabling the identification of resources impacting response performance through correlation management and data migration processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If all resources associated with the monitored resource are output to the administrator based on mapping information, then comprehensive performance information is provided, but the administrator cannot promptly and accurately identify the truly problematic resource
Solution Approach 1:
The patent extracts and presents only the specific resources that are actually affecting the performance of the monitored resource, rather than presenting all associated resources. This is achieved by analyzing I/O request status and resource relationships to identify and extract only the problematic resources that need administrator attention.
Solution Approach 2:
The patent applies local quality by providing different information presentation for different resources based on their actual impact on performance. Resources that are truly affecting performance are highlighted and presented with detailed information, while resources that are not affecting performance are excluded or minimized in the presentation.
2Loss of time
If all resources are presented to the administrator, then complete resource status information is available, but the administrator cannot promptly identify the bottleneck causing performance deterioration
Solution Approach 1:
The patent extracts only the critical information about resources that are actually causing performance issues, removing unnecessary information about resources that are not affecting performance. This extraction process significantly reduces the time required for administrators to identify problematic resources.
Solution Approach 2:
The patent segments the resource information presentation into different categories based on their impact on performance. Critical resources that are affecting performance are segmented and presented separately with prioritized information, while non-critical resources are either excluded or presented in a condensed format.
3Loss of information
If resources that do not receive I/O requests are included in the presentation, then complete resource mapping is shown, but the information becomes less relevant to actual performance issues
Solution Approach 1:
The patent extracts and excludes resources that do not receive I/O requests or are not affecting performance from the presentation to the administrator. By analyzing I/O request status and resource utilization, the system identifies and extracts only the relevant resources that contribute to performance issues, removing irrelevant information.
Solution Approach 2:
The patent applies local quality by adjusting the information presentation based on the actual performance impact of each resource. Resources with high I/O activity and performance impact are presented with detailed information, while resources with no or minimal I/O activity are excluded or presented with minimal information.
Data Source
AI summary
A computer system is provided which includes a host computer having a volume, a storage apparatus including a storage media, the storage apparatus providing a RAID group configured by the storage media, a logical volume configured by the RAID group, a pool configured by the logical volume, and a virtual volume allocated from the pool and corresponding to the volume of the host computer. The computer system also includes a management server to store performance information regarding the logical volume and/or RAID group and mapping information regarding a relationship among components of the computer system in view of logical and physical connections. The management server determines whether a correspondence relationship exists among the components by referring to the performance information, and outputs a relationship among the components based on the determination of the correspondence relationship and the mapping information.