Storage Relay Device Aggregates Error Data to Reduce Mirror Transfer Latency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In storage systems, maintaining data redundancy through mirror transfer between controller modules (CMs) is challenging due to potential internal abnormalities in devices, which can lead to system failures and require time-consuming path disconnection for error handling, affecting delivery guarantee and IOPS performance.
Innovation Solution
A storage system with a relay device that detects abnormalities, aggregates error information, and uses a delivery guarantee function to determine if mirror transfer is complete, allowing for immediate error reporting and minimizing path disconnection, thereby reducing latency and improving IOPS performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If path disconnection is performed for error handling in mirror transfer, then system reliability is improved, but delivery guarantee time increases and IOPS performance deteriorates
Solution Approach 1:
The patent applies preliminary action by having the relay device detect and aggregate error information from multiple devices before mirror transfer completion. The error information is pre-aggregated in a register, so when delivery guarantee checking is needed, the source CM can immediately read the aggregated error state without waiting for individual device error reports, thus reducing delivery guarantee time while maintaining reliability
Solution Approach 2:
The patent uses an intermediary approach by introducing a relay device that acts as a mediator between multiple CMs and the source CM. The relay device aggregates error information from various devices in the mirror transfer path and consolidates it in a single register, serving as an intermediary that simplifies error checking for the source CM and reduces the time needed for delivery guarantee verification
2Measurement precision
If individual device error detection is performed without aggregation, then error detection precision is maintained, but device complexity increases and processing efficiency decreases
Solution Approach 1:
The patent applies merging by combining error detection and aggregation functions into a single relay device. Multiple devices' error information is merged into one centralized register at the relay device, reducing the complexity that would otherwise exist at each individual device while maintaining precise error detection through the aggregated view
Data Source
AI summary
A storage system includes a plurality of controllers and a relay device and is configured to perform mirror transfer between the plurality of controllers, the relay device detects an abnormality in each device located between the plurality of controller and aggregates error information in a register, and a first controller module that is a source of performing the mirror transfer among the plurality of controllers reads content of the register and determines whether the mirror transfer is completed normally, after the mirror transfer is performed.