Proactive Storage Resource Release via Load Score Demotion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Previous data storage systems fail to effectively deallocate high performance storage resources once a storage group meets its service level objective, leading to depletion of available high performance storage and inefficient use of resources as data access patterns change, causing increased storage costs.
Innovation Solution
The system calculates a load score for each unit of storage allocated from a high performance disk group, determines a demotion threshold, and migrates data from high performance to lower performance storage when the service level objective is exceeded by a safety margin, allowing for proactive deallocation and reallocation of resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If high performance storage resources are allocated to storage groups to meet service level objectives, then the service level objective compliance is improved, but the availability of high performance storage resources for other storage groups deteriorates
Solution Approach 1:
The patent implements dynamic allocation and deallocation of high performance storage resources based on real-time monitoring of service level objective compliance. When a storage group exceeds its SLO by a safety margin, the system proactively deallocates high performance storage resources; when the SLO is met, resources are reallocated. This dynamic adjustment resolves the contradiction by making high performance storage availability variable rather than static, allowing resources to be released for other uses while maintaining SLO compliance.
Solution Approach 2:
The patent discards (deallocates) high performance storage resources from storage groups that are exceeding their service level objectives, and recovers these resources for reallocation to other storage groups that may need them. This cyclic process of discarding and recovering resources resolves the contradiction by ensuring that high performance storage is continuously available to those storage groups that actually need it to meet their SLOs, rather than being permanently allocated to groups that may not need it at all times.
2Reliability
If high performance storage resources are continuously allocated to storage groups, then the service level objective is maintained, but storage costs increase due to inefficient resource utilization
Solution Approach 1:
The system dynamically adjusts high performance storage allocation based on actual SLO compliance needs. Instead of continuous allocation, the system monitors performance metrics and only maintains high performance allocation when the storage group is exceeding its service level objective by a safety margin. When the SLO is met, the system deallocates resources, reducing storage costs while maintaining reliability only when necessary.
Solution Approach 2:
The patent changes the allocation parameter (amount of high performance storage) based on the service level objective compliance status. The system transitions between different allocation states: full allocation when SLO is exceeded, partial or no allocation when SLO is met. This parameter change resolves the contradiction by optimizing the balance between reliability maintenance and cost reduction.
3Reliability
If data is migrated to high performance storage to reduce response time, then the service level objective compliance is improved, but the response time for deallocating and remigrating data increases
Solution Approach 1:
The system performs preliminary actions by proactively monitoring service level objective compliance and initiating deallocation before resources are completely depleted or performance critically degrades. The safety margin threshold triggers early deallocation, allowing the system to respond to changing data access patterns without emergency migrations. This preliminary action reduces the urgency and time required for data migration operations.
Solution Approach 2:
The patent implements a feedback mechanism where the system continuously monitors service level objective compliance and uses this information to control allocation and deallocation decisions. When the SLO is met, feedback triggers deallocation; when the SLO is exceeded, feedback triggers reallocation. This closed-loop feedback system resolves the contradiction by ensuring timely responses to performance changes, minimizing unnecessary migration time while maintaining compliance.
Data Source
AI summary
High performance storage resources are proactively deallocated when a storage group exceeds its service level objective. A load score is calculated for each unit of storage allocated from a high performance physical disk group. A minimum load score is determined that is the lowest load score calculated for the storage units allocated from the high performance disk group to the storage group. When the storage group service level objective is exceeded, a demotion threshold is calculated equal to a product of the minimum load score and a demotion range factor. A demotion set is determined made up of storage units allocated from the high performance disk group having load scores less than or equal to the demotion threshold. Host data stored on the demotion set is migrated to storage from a lower performance disk group, and the high performance units of storage in the demotion set are released for re-allocation.