Runtime Command Selection for Serial Data Restore
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Accessing non-contiguous data on serial access resources, such as data files or object stores, is inefficient due to high latency and resource allocation times for commands like open, seek, and close, leading to slow restore operations during partial data restoration.
Innovation Solution
A system that selects optimal combinations of access commands at runtime based on section locations and performance targets, such as minimizing bandwidth, time, or cost, to optimize data access by generating and executing combinations of open, seek, read, and close commands efficiently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If traditional access commands (open, seek, close) are used to access non-contiguous data on serial access resources, then data can be accessed, but restore time increases and performance decreases
Solution Approach 1:
The patent segments the access to non-contiguous data into multiple parallel operations. Instead of sequentially executing open-seek-close cycles for each data section, the system divides the backup file into multiple sections and accesses them simultaneously using separate file handles, thereby reducing the total restore time and improving data access efficiency.
2Reliability
If multiple open, seek, and close commands are executed for non-contiguous data access, then complete data retrieval is achieved, but resource allocation time and latency increase
Solution Approach 1:
The patent performs preliminary actions by opening multiple file handles in advance before actually reading the data sections. This allows the system to prepare all necessary access paths beforehand, reducing the resource allocation time during the actual data retrieval process while ensuring complete data retrieval through multiple parallel handles.
3Ease of operation
If sequential access commands are used for non-contiguous data, then data access is possible, but restore speed decreases
Solution Approach 1:
The patent segments the restore operation into multiple parallel data stream operations. By dividing the backup file into multiple sections and accessing them simultaneously through separate handles, the system maintains operational simplicity while dramatically improving restore speed compared to sequential access methods.
Solution Approach 2:
The patent merges multiple parallel data stream operations into a single restore process. By combining multiple simultaneous read operations from different sections of the backup file into one coordinated restore operation, the system achieves both ease of operation and high restore speed.
Data Source
AI summary
Run-time selection of optimal commands for accessing serial access resources is described. A system receives a request to access a serial access resource. The system identifies a first section to access in the serial access resource and a second section to access in the serial access resource which is separated from the first section to access by an intermediate section in the serial access resource. The system generates combinations of access commands associated with accessing the first and the second sections to access. The system estimates target values, based on locations of the first and the second sections to access in the serial access resource, corresponding to the combinations of access commands. The system selects a combination of access commands which corresponds to an optimal target value. The system enables performance of the selected combination of access commands on the first and the second sections in the serial access resource.