Hierarchical Storage Tier Migration Based on Access Frequency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data in superior storage tiers is wastefully consumed as it remains arranged even when the application is not activated or after activation, leading to inefficient use of high-performance storage resources.
Innovation Solution
A method for controlling hierarchical storage that involves monitoring access patterns and migrating data between storage tiers based on access frequency, ensuring that only actively needed data is stored in high-performance tiers, while less accessed data is moved to lower tiers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data is permanently arranged in superior storage tiers to satisfy response performance requirements, then application response performance is improved, but storage resource efficiency deteriorates due to wasteful consumption of high-performance storage capacity
Solution Approach 1:
The patent applies dynamics by making the storage tier arrangement dynamic rather than static. Data is automatically migrated between superior and inferior storage tiers based on real-time access frequency monitoring. When access frequency exceeds a threshold, data is migrated to superior tiers; when it falls below, data is migrated to inferior tiers. This dynamic adjustment ensures that only actively accessed data occupies high-performance storage resources, resolving the contradiction between maintaining response performance and improving storage efficiency.
2Speed
If data is arranged in superior storage tiers to meet demand during peak usage, then service response time is improved, but storage cost increases due to over-provisioning of high-performance capacity
Solution Approach 1:
The patent applies parameter changes by monitoring access frequency as a dynamic parameter and using it to determine data placement decisions. The system changes the storage tier parameter (from superior to inferior or vice versa) based on whether access frequency exceeds a predetermined threshold. This parameter-driven approach allows the system to optimize service response time during peak usage while avoiding the cost of permanently over-provisioning high-performance storage capacity for all data.
3Loss of time
If all data is kept in superior storage tiers to ensure immediate access, then application activation performance is improved, but the superior tier capacity is quickly exhausted
Solution Approach 1:
The patent applies preliminary action by proactively migrating data to superior storage tiers before it is needed, based on predicted access patterns. When access frequency exceeds the threshold, the system pre-migrates data to superior tiers in advance, ensuring that when the application is activated, the data is already in the high-performance storage location. This eliminates activation delays without requiring all data to permanently reside in superior tiers, thus preserving capacity.
Data Source
AI summary
A method for controlling hierarchical storage including: a first step for storing first information relating to the association between the specific processes and the storage regions of the storage tiers; a second step for obtaining second information relating to the access operations of the specific processes as a function of time; a third step for obtaining third information relating to the amount of access to the storage regions as a function of time; and a fourth step for identifying the time of occurrence of a change in the amount of access, from the second information and the third information on the basis of the first information, and determining, according to the identified time of occurrence of the change in the amount of access, a transfer initiation time at which data in the storage regions is to be transferred between storage tiers.