Storage Apparatus Silent Data Corruption Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Advanced storage apparatus architectures and multifunctional native flashes complicate storage control microprograms, leading to increased risks of silent data corruption and incorrect data restoration due to broken addresses in mapping tables, where conventional error detection methods fail to accurately identify and correct data errors, especially when both parity codes are broken.
Innovation Solution
A storage apparatus with a control unit that manages data elements across multiple storage devices, using multiple redundancy codes generated by different calculation methods to identify and correct incorrect data elements by comparing restoration calculations from these codes, ensuring accurate data restoration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional error detection methods using single parity code are used, then the system is simple to operate, but the reliability of error detection deteriorates when both parity codes are broken
Solution Approach 1:
The patent applies local quality by making different parts of the redundancy system have different properties. Specifically, it uses multiple parity codes (first parity and second parity) with different calculation methods stored at different locations, allowing the system to detect and correct errors even when one parity code is corrupted by comparing results from different calculation approaches
Solution Approach 2:
The patent employs composite materials principle by combining multiple redundancy codes generated through different calculation methods. The control unit stores both first parity and second parity codes, and when error correction is needed, it calculates correction values using both parity codes and selects the appropriate correction based on comparison, creating a composite error detection system that is more robust than single parity code
2Adaptability or versatility
If storage control microprograms are enhanced with advanced architectures and multifunctional native flashes, then the functionality and capacity efficiency are improved, but the risk of silent data corruption increases
Solution Approach 1:
The patent applies preliminary action by pre-calculating and storing multiple parity codes (first parity and second parity) along with the data before any corruption occurs. These redundancy codes are stored in advance using different calculation methods, so when silent data corruption happens, the system already has multiple correction paths available without needing to perform complex real-time analysis
Solution Approach 2:
The patent implements feedback mechanism where the control unit calculates correction values using both first parity and second parity codes, then compares these correction values to determine which one to apply. This feedback loop ensures that even when one parity code is corrupted, the system can detect the discrepancy and use the correct parity code to restore data integrity
Data Source
AI summary
The storage apparatus has a control unit that includes: an identification unit that is configured to determine that a first data element contained in the data is incorrect, when a first restoration calculation first data element, which is restored from the first data element using other data elements of the data excluding the first data element and a new redundancy code obtained from the data by a first calculation method, coincides with a second restoration calculation first data element, which is restored from the first data element using the other data elements and a new redundancy code obtained from the data by a second calculation method; and a restoration unit that is configured to correct the first data element in the storage devices that is determined to be incorrect by the identification unit, to either the first restoration calculation first data element or the second restoration calculation first data element.