Storage Control Device for Collective Block Reading
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Distributed storage systems using hard disk drives (HDDs) or tape devices face inefficiencies in reading data due to the need to read and combine multiple blocks from different storage devices, leading to increased reading times, especially when using tape devices where frequent movement of the tape head results in longer data retrieval times.
Innovation Solution
A storage control device that divides data into original and copy blocks and distributes them across multiple storage devices, allowing for collective reading of these blocks from the most efficient storage device, thereby reducing the number of locations the tape head needs to access and improving reading speed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data blocks are distributed across multiple storage devices, then data redundancy and reliability are improved, but reading time increases due to needing to read and combine blocks from multiple devices
Solution Approach 1:
The patent applies preliminary action by pre-grouping related data blocks (original blocks and copy blocks) into the same storage device during the data writing phase. This advance organization ensures that when read requests occur, all necessary blocks are already co-located, eliminating the need for cross-device reading and combining operations.
Solution Approach 2:
The patent merges the storage of original blocks and copy blocks into the same storage device, creating a unified storage unit. This merging allows the storage system to treat grouped blocks as a single logical unit, enabling efficient collective reading operations from one device rather than multiple separate read operations.
2Quantity of substance
If tape devices are used for storage, then storage capacity and cost-effectiveness are improved, but reading speed deteriorates due to frequent tape head movement
Solution Approach 1:
The patent applies preliminary action by pre-organizing all required data blocks (original and copy blocks) to be stored on the same tape device before reading operations. This advance grouping ensures that tape heads can service multiple related blocks in a single pass without frequent repositioning, significantly reducing seek time and improving reading speed while maintaining tape storage's capacity advantages.
Solution Approach 2:
The patent enables continuous useful action by ensuring that all blocks needed for a given object are located on the same tape device, allowing the tape drive to read data continuously without interruption for head movement. This eliminates idle time during tape operations and maximizes the utilization of the tape drive's sequential access capability.
3Reliability
If blocks are read from multiple storage devices, then data redundancy is maintained, but the number of tape head locations increases leading to longer retrieval times
Solution Approach 1:
The patent merges the storage locations of original blocks and copy blocks into a single storage device, reducing the number of tape head locations from multiple devices to one. This consolidation maintains data redundancy (since both original and copy blocks are present) while dramatically simplifying the physical access pattern and reducing retrieval time.
Solution Approach 2:
The patent applies segmentation by dividing the storage system into logical groups where each storage device contains a complete set of related blocks (original and copy). This segmentation strategy organizes data at the block group level rather than distributing individual blocks across devices, reducing the complexity of tape head movements while preserving redundancy through the original-copy block structure within each device.
Data Source
AI summary
A storage control device, includes a processor configured to: dispose original blocks to which data on an object basis is divided and assigned and copy blocks which are copies of the original blocks respectively alternately in storage devices of a distributed storage system in a distributed manner; determine, upon reception of a read request for an object, as a first storage control device out of the storage control devices a read task which collectively reads the original blocks and the copy blocks which belong to the object from one of the storage devices; notify the storage control devices of the read task; upon reception of the read request as the first storage control device, collectively read the original blocks and the copy blocks; and transfer the data of the blocks which have been read to a second storage control device which is the read task for the blocks.