Sub-LUN Data Segment Tiering via Timing Parameters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional data storage tiering methods lack application-specific, time-based mechanisms for managing sub-LUN data segments, leading to inefficient data placement and delayed migration of data to high-performance tiers, which can impact performance in applications requiring immediate access, such as financial reporting.
Innovation Solution
An API is provided for host applications to configure sub-LUN data segment tiering preferences with timing parameters, allowing for application-specific tiering that overrides conventional tiering methods, enabling data to be relocated to preferred tiers during specific time periods, such as peak performance needs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If conventional data storage tiering methods are used, then data is automatically placed on storage tiers based on general access patterns, but data migration to high-performance tiers is delayed and lacks application-specific timing control
Solution Approach 1:
The system performs preliminary actions by pre-scheduling data migration to high-performance tiers before critical time periods arrive. The storage system receives timing information from applications and proactively migrates data segments to appropriate tiers in advance, ensuring data is ready when needed without waiting for conventional on-demand migration triggers.
Solution Approach 2:
The invention changes the timing parameter of data migration from passive, event-driven conventional tiering to active, application-specified time-based tiering. By incorporating timing parameters into the tiering decision process, the system can control when data is migrated based on application-specific time information, transforming the migration behavior from generic to precisely timed.
2Productivity
If data is placed on high-performance tiers continuously, then application performance is improved, but storage system costs increase due to inefficient use of expensive storage resources
Solution Approach 1:
The system implements periodic action by migrating data to high-performance tiers only during specific time periods when applications require peak performance. Instead of continuous placement, data is dynamically moved to and from high-performance tiers based on time-based policies, ensuring expensive storage resources are utilized only when necessary for critical operations.
Solution Approach 2:
The invention introduces dynamics to data placement by making tier assignment flexible and time-dependent rather than static. Data segments can be automatically promoted to or demoted from high-performance tiers based on changing time conditions and application requirements, optimizing the balance between performance needs and resource efficiency.
3Ease of operation
If conventional LUN-level tiering is used, then data placement is simplified, but fine-grained control over specific data segments is lost
Solution Approach 1:
The system applies segmentation by dividing LUNs into smaller, independently manageable data segments or extents. This allows fine-grained control over specific portions of data, enabling different tiering policies to be applied to different segments within the same LUN. The segmentation enables precise control over which data segments are migrated and when, while maintaining manageable complexity through automated policies.
Data Source
AI summary
A technique for managing temporal data placement in data storage systems is disclosed. An application program interface (API) is provided, where a host application is configured to communicate sub-LUN data segment location, application configuration tiering, and timing parameter information to the API. The sub-LUN data segment location, application configuration tiering, and timing parameter information is transmitted to the data storage system via an out-of-band communications path. The sub-LUN data segment location, application configuration tiering, and timing parameter information is received at the data storage system and associated with the corresponding LUN. A timing estimate for implementing sub-LUN data segment relocation is generated based on system derived tiering information and recent relocation timing information. Relocation of the sub-LUN data segment is scheduled such that the sub-LUN data segment is relocated to a tier corresponding to the application configuration tiering information during the time period associated with the timing parameter.