Hybrid SSD Disk File Delivery Startup Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing file delivery systems face challenges in providing fast startup and efficient delivery of large files due to slower read times from disk drives compared to solid state drives, especially when handling multiple simultaneous requests.
Innovation Solution
The system optimizes file delivery by writing initial blocks of files to both disk and solid state drives, specifying block sizes based on file size and network bandwidth, and utilizing an interface module to cache and manage data blocks for rapid transfer, ensuring that subsequent blocks are ready for immediate delivery without pauses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If files are stored only on disk drives, then storage capacity is maximized, but read speed is slow causing long startup times
Solution Approach 1:
The storage system is segmented into two distinct parts: a solid state drive (SSD) for storing frequently accessed initial blocks of files, and a disk drive for storing the remaining file data. This segmentation allows the system to leverage the high read speed of SSDs for critical initial data while maintaining the large capacity of disk drives for comprehensive file storage.
Solution Approach 2:
Different storage media are assigned different functional qualities within the same storage system. The SSD provides fast read access for initial blocks, while the disk drive provides economical storage for complete files. This local quality differentiation optimizes both speed and capacity based on data access patterns.
2Loss of time
If initial blocks are written to both disk and solid state drives, then startup time is reduced, but storage space is consumed
Solution Approach 1:
The initial blocks of files are pre-written to both the SSD and disk drive during the file storage process. This preliminary action ensures that when a user requests a file, the initial blocks are already available on the high-speed SSD, enabling rapid startup without waiting for sequential disk reads.
Solution Approach 2:
Instead of duplicating entire files on both storage media (excessive action), only the initial blocks are replicated (partial action). This partial duplication provides sufficient speed boost for startup while minimizing the storage space consumed by avoiding redundant storage of complete file contents.
3Productivity
If block sizes are optimized for network transfer, then transfer rate is maximized, but read time from disk increases
Solution Approach 1:
Initial blocks are pre-loaded onto the SSD with optimized block sizes for network transfer, so that when file delivery begins, the data is already positioned for rapid retrieval. This preliminary preparation eliminates the read time penalty that would otherwise occur when reading from the disk drive during the critical initial transfer phase.
Data Source
AI summary
A file delivery system includes a disk file system, a solid state drive file system, and an interface module. The disk file system includes a disk drive that stores a plurality of files for download by a user device. The solid state drive file system includes a solid state drive that stores a first block of data for each of the files stored on the disk drive. The interface module is in communication with the disk file system and with the solid state drive file system, and is configured to retrieve the first block of data of one of the files from the solid state drive in response to a request from the user device for one of the files, to send a read command to the disk drive for a next block of the one of the files at substantially the same time as the first block is retrieved from the solid state drive and sent to the user device, and to store the next block in a buffer.