SCSI Target Link Anomaly Tolerance Mechanism
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
SCSI systems face challenges in robustness, particularly in tolerating communication anomalies and command bursts, which can lead to futile retries and increased recovery efforts due to limited retry capabilities and self-compounding command bursts.
Innovation Solution
Implementing a method where the SCSI target withholds responding to failed commands until communication links are determined to be normal or a predetermined timeout expires, and proactively returning responses for pending commands within an initiator-defined timeout duration to prevent SCSI initiator timeouts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the SCSI target immediately returns a response for a failed command, then the response time is reduced, but futile retries occur when communication links are temporarily abnormal
Solution Approach 1:
The SCSI target performs preliminary checking of communication link status before returning a response for a failed command. By proactively verifying link normality in advance, the system determines whether to suppress the response, thereby avoiding futile retries while maintaining efficient command processing when links are operational.
2Reliability
If the SCSI initiator increases retry attempts to handle communication failures, then link anomaly tolerance is improved, but command burst tolerance deteriorates due to self-compounding command bursts
Solution Approach 1:
The SCSI target actively monitors communication link status and provides feedback by suppressing responses when link abnormalities are detected. This feedback mechanism prevents the SCSI initiator from performing unnecessary retries, thereby reducing command burst generation and improving overall system productivity during temporary link failures.
3Productivity
If the SCSI target suppresses responses for failed commands during link abnormalities, then futile retries are reduced, but the complexity of determining link status increases
Solution Approach 1:
The SCSI target utilizes its existing communication infrastructure and error detection mechanisms to self-determine link status. By leveraging built-in capabilities such as checking for command transmission acknowledgments and utilizing existing error detection fields, the system avoids the need for separate complex link monitoring subsystems.
Data Source
AI summary
A computer-implemented method is disclosed to enhance link anomaly tolerance in a small computer system interface (SCSI) system. The method starts with detecting a SCSI command failed at a SCSI target, where the SCSI target is communicatively coupled with a SCSI initiator through a set of communication links and the failed SCSI command is sent from the SCSI initiator. After detecting the failure, the SCSI target withholds returning a response for the failure. Then the SCSI target checks a status of the set of communication links periodically within the withholding duration while the response for the failed SCSI command is withheld. The SCSI target returns the response for the failed SCSI command upon at least one of two conditions is met: (1) The withhold duration expires; (2) The set of communication links is determined to function normally based on checking the status of the set of communication links.