Storage Capacity Planning Using Predictive Load Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In storage systems with mixed media of different performances, determining the appropriate amount of storage capacity to install or remove to optimize performance is challenging, leading to potential over-installment or under-installment, which can degrade performance or increase costs unnecessarily.
Innovation Solution
A storage apparatus with virtual volumes and a controller that manages actual area groups by estimating future capacity needs based on past access patterns and load thresholds, allowing for precise calculation and adjustment of installed or removed capacity to match performance requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If storage capacity is increased to improve performance, then storage capacity is improved, but cost increases and performance may be over-improved
Solution Approach 1:
The storage pool is segmented into multiple media types (e.g., SSD, SAS, SATA) with different performance characteristics. Each media type is managed separately with individual capacity planning, allowing precise allocation of storage capacity to match actual performance requirements rather than uniformly increasing all storage capacity
Solution Approach 2:
The system changes the parameter of storage capacity planning from static to dynamic by using predictive algorithms that estimate future capacity needs based on historical growth rates. This allows capacity to be increased only when and where actually needed, avoiding unnecessary cost increases
2Quantity of substance
If storage capacity is increased to improve performance, then storage capacity is improved, but performance may be over-improved
Solution Approach 1:
The system implements feedback mechanisms by continuously monitoring actual storage usage patterns and performance metrics, then using this information to adjust capacity planning decisions. The predictive algorithm compares estimated future needs with actual growth rates, allowing the system to avoid over-provisioning storage capacity that would not be utilized
Solution Approach 2:
The capacity planning system transitions from static to dynamic by enabling capacity adjustments at granular levels (per media type, per LUN, per volume) based on real-time and predictive data. This dynamic approach ensures storage capacity is optimized to match actual performance requirements rather than being uniformly increased
3Ease of operation
If storage capacity planning is performed without considering media performance differences, then planning simplicity is improved, but planning precision deteriorates
Solution Approach 1:
The system applies local quality by treating different media types (SSD, SAS, SATA) as distinct entities with different performance characteristics and capacity planning requirements. Each media type receives customized capacity planning based on its specific properties and the performance requirements of workloads assigned to it, rather than applying a uniform planning approach
4Ease of manufacture
If storage capacity is removed to reduce cost, then cost is reduced, but performance may be degraded
Solution Approach 1:
The system performs preliminary actions by proactively identifying storage capacity that can be removed through predictive analysis of future capacity needs. By estimating capacity requirements in advance and comparing them with current allocations, the system can safely remove excess capacity before it becomes wasteful, ensuring performance reliability is maintained while reducing costs
Data Source
AI summary
A storage apparatus coupled to a host device comprises a virtual volume which is a virtual logical volume configured of multiple virtual areas and a pool configured of multiple actual area groups of different performances. A controller manages pool status information which is the information showing which actual area is allocated to which virtual area and access load related to the virtual areas. A management system of the storage apparatus, with reference to the pool status information at multiple points of time from past to present and an access load threshold which is equal to or larger than 1, estimates the used capacity of each actual area group at points of time in the future, calculates the installed/removed amount of each actual area group which is the difference between the estimated used capacity and the current storage capacity, and performs the processing based on the calculated result.