Database Storage Zone Allocation by Access Frequency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In massively parallel processing systems, existing data allocation mechanisms based on least disk utilization can lead to undesirable system performance due to node skew, as they do not effectively account for varying access frequencies and storage device zone speeds.
Innovation Solution
A device profiler evaluates storage device zones for speed quality ratings and a temperature monitor updates data access frequency quantifiers, allowing data to be allocated and migrated to zones with matching or proximate speed quality ratings based on access frequency, optimizing storage and retrieval.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If least disk utilization algorithm is used for data allocation, then storage space distribution is simplified, but system performance deteriorates due to node skew and failure to account for access frequencies
Solution Approach 1:
The patent divides storage devices into multiple zones with different performance characteristics (inner zones with higher access speeds, outer zones with lower access speeds). Data is allocated to specific zones based on access frequency requirements, matching high-frequency data to fast zones and low-frequency data to slower zones, thereby optimizing overall system performance while maintaining manageable complexity through structured zone management.
Solution Approach 2:
The patent introduces data temperature as a dynamic parameter to characterize access frequency and uses this parameter to determine optimal storage zone allocation. By continuously monitoring and updating data temperature, the system adapts data placement decisions to current access patterns, resolving the contradiction between simple allocation rules and performance optimization.
2Speed
If data is stored on faster storage zones, then access speed for frequently accessed data improves, but storage device complexity and migration overhead increase
Solution Approach 1:
The patent implements dynamic data placement by continuously monitoring data temperature (access frequency) and automatically migrating data between storage zones based on current access patterns. This dynamic approach ensures that data is always stored in zones appropriate for its current access frequency, optimizing access speed while the automated migration process manages the complexity of maintaining optimal placements.
Solution Approach 2:
The system uses data temperature as a feedback mechanism to guide data migration decisions. By monitoring access frequencies and using this information to determine when and where to migrate data, the system automatically adjusts storage placements to optimize access speed without requiring complex manual management, as the feedback loop handles the complexity of coordinating migrations across the distributed storage system.
Data Source
AI summary
The frequency with which data is accessed within the system may be periodically monitored and a corresponding access frequency quantifier assigned to the data is updated accordingly. The data access frequency quantifier may be associated with a storage device zone speed quality rating. The association between data access frequency quantifiers and the storage device zone speed quality ratings may be made in a hierarchical association such that quantifiable differentials may be ascertained between a particular access frequency quantifier and a storage device zone speed quality rating. In this manner, when no storage zone having a speed quality rating that is associated with data having a particular access frequency quantifier is available for storage of the data, a storage zone having a speed quality rating more proximate the speed quality rating associated with the access frequency quantifier may be identified for migration.