Disk-Based Sort Procedure for Sequential Block Storage Updates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional storage systems face inefficiencies in performing both random write operations and sequential read operations, with conventional disk arrays providing good sequential read performance but poor random write performance, and large write buffers being costly to maintain.
Innovation Solution
A disk-based sort procedure is implemented to establish locality among updates in a disk-based log, allowing write data to be efficiently written to home locations on a home location array, reducing resource consumption and improving disk update performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If a conventional disk array employs a static layout that maps sequential addresses into sequential locations on disk, then sequential read performance is improved, but random write performance deteriorates
Solution Approach 1:
The patent implements a dynamic mapping approach where the disk array transitions from a static address-to-location mapping to a dynamic one. The log-structured file system dynamically determines write locations based on current disk state, allowing the system to adapt to varying workload patterns and maintain both sequential read and random write performance.
Solution Approach 2:
The patent introduces a log-structured file system as an intermediary layer between the host system and the physical disk array. This intermediary translates random write requests into sequential disk operations by buffering writes in a log structure, thereby resolving the conflict between random write performance and sequential read performance.
2Productivity
If a conventional log-structured file system converts random write operations into sequential disk transfers, then random write performance is improved, but sequential read performance deteriorates due to randomized block locations
Solution Approach 1:
The patent performs preliminary organization of data blocks during the write operation. By pre-calculating and pre-organizing block locations in the log structure before actual disk writing occurs, the system ensures that subsequent sequential read operations can efficiently access data without suffering from randomized locations.
Solution Approach 2:
The log-structured file system acts as an intermediary that maintains mapping information between logical block addresses and physical disk locations. This intermediary layer enables the system to translate sequential read requests into efficient physical access patterns while maintaining the benefits of log-structured writing.
3Productivity
If large write buffers are implemented in non-volatile memory to compensate for poor write performance, then write performance is improved, but system cost and complexity increase
Solution Approach 1:
The patent makes the existing disk array serve multiple functions: it acts as both the primary storage medium and the write buffer. By utilizing the disk array's own capacity for buffering writes through the log-structured file system, the system eliminates the need for separate non-volatile memory buffers, thereby reducing complexity while maintaining write performance.
Solution Approach 2:
The disk array serves itself as the buffer by using its own storage capacity through the log-structured file system. The system buffers writes internally within the disk array rather than requiring external buffer memory, allowing the disk array to compensate for its own performance limitations without additional components.
Data Source
AI summary
An improved system and method enhances performance of updates to sequential block storage of a storage system. A disk-based sort procedure is provided to establish locality among updates (write data) held in a disk-based log, thereby enabling the write data to be efficiently written to home locations on a home location array. As the write data is received, a log manager of the storage system temporarily stores the data efficiently on the disk-based log. As more write data arrives, the log manager sorts the data in the log in accordance with the sort procedure, thus increasing the locality of data when stored on the home location array. When the log approaches capacity, the log manager writes the sorted data to their home locations on the array with high locality and performance.