Storage Controller Timeout Management for Response Time
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing storage control apparatuses experience deteriorated response performance and service quality due to prolonged response times when retry processing is performed on hard disks, particularly affecting applications that require rapid data access, such as ticketing and video distribution programs.
Innovation Solution
The storage control apparatus sets a shorter timeout time for data retrieval from storage apparatuses based on specific conditions, such as queuing mode and load status, and employs a second management unit to manage timeout errors, enabling data generation and transfer from alternative storage apparatuses to maintain response time guarantees.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If retry processing is performed on hard disk when data cannot be read, then data reliability is improved, but response time becomes longer and service quality deteriorates
Solution Approach 1:
The patent segments error handling into two distinct pathways: timeout errors (managed by second management unit) and other failures (managed by first management unit). This segmentation allows differential processing where timeout errors trigger immediate correction copy without retry, while other failures undergo traditional retry processing, thus resolving the contradiction between reliability and response time.
Solution Approach 2:
The patent changes the timeout parameter of the hard disk to a shorter value when timeout errors are detected. This parameter change prevents prolonged waiting for unresponsive disks, thereby reducing response time while maintaining data reliability through alternative correction mechanisms.
2Reliability
If timeout time is set to first value (longer time) for data retrieval, then data reading completeness is improved, but response performance deteriorates under high load
Solution Approach 1:
The patent dynamically adjusts the timeout time parameter based on the type of error detected. For timeout errors, it uses a shorter timeout (second value), while for other failures, it uses the longer timeout (first value). This dynamic adjustment optimizes response performance under high load while maintaining data reading completeness when appropriate.
Solution Approach 2:
The patent changes the timeout parameter from a fixed value to a conditional value that switches between first value and second value based on error type. This parameter change enables the system to adapt to different operational conditions, improving response performance without sacrificing data integrity.
3Reliability
If correction copy is performed to restore data from other hard disks, then data reliability is improved, but processing time increases
Solution Approach 1:
The patent performs correction copy in advance for timeout errors detected in the first management unit, storing the restored data before it is needed. This preliminary action ensures that when data is required, it is already available, eliminating the need for time-consuming correction operations at the moment of data access.
Solution Approach 2:
The patent prepares correction copy data beforehand and stores it in a cushion area, creating a buffer that prevents future processing delays. This beforehand cushioning ensures that even if the original disk fails, the restored data is already available, thus maintaining both reliability and fast response time.
Data Source
AI summary
[This invention] inhibits the response time of the storage control apparatus from being longer even if the response time of the storage apparatus is long.The disk adapter (DKA), receiving a read message from the channel adapter (CHA), sets the timeout time in accordance with specified conditions, and tries to read data from the storage apparatus 4. As the timeout time, either the normal value or the shortened value is selected. If a timeout error occurs, the read job is reset, and correction read is started.