Silent Data Corruption Protection in Clustered Virtual Environments
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems fail to prevent the propagation of undetectable silent data corruption in clustered virtual environments, which can lead to cascading failures and data integrity issues due to errors caused by factors like cosmic radiation and unreliable power supplies.
Innovation Solution
A method involving a processor that performs a data integrity check and compares the results with a secondary system's integrity check, only allowing data propagation if the results match, thereby preventing silent data corruption from spreading throughout the network.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data are propagated directly to the network environment without verification, then system productivity is maintained, but data integrity deteriorates due to silent data corruption
Solution Approach 1:
The patent performs data integrity checks (CRC calculations) on data before propagating it to the network environment. The primary computer system calculates a first integrity check result, and a secondary system calculates a second integrity check result, comparing both results to detect silent data corruption before the corrupted data can propagate through the network, thus preventing cascading failures while maintaining system productivity
Solution Approach 2:
The patent introduces an intermediary verification mechanism where a secondary computer system acts as a mediator to independently verify the integrity of data prepared by the primary system. This intermediary check provides an additional layer of validation without significantly impacting the primary data flow, resolving the contradiction between reliability and productivity
2Reliability
If a secondary redundant component is deployed to mirror processes, then system reliability is improved for detectable errors, but the system remains vulnerable to silent data corruption propagation
Solution Approach 1:
The patent implements a feedback mechanism where integrity check results from both primary and secondary systems are compared. When a mismatch is detected, the system identifies potential silent data corruption and prevents the corrupted data from propagating further, providing continuous monitoring and correction capability that addresses the limitation of traditional redundancy
Solution Approach 2:
The patent uses the secondary computer system as a copy verification mechanism. Instead of relying solely on process mirroring, the secondary system independently calculates integrity checks on copied data, providing a verification layer that can detect silent corruptions that process mirroring alone cannot detect
Data Source
AI summary
In one embodiment, a computer program product includes a computer readable medium having stored thereon computer readable program instructions. The instructions are configured to cause modules of a computer system to: determine whether data are in condition to be propagated beyond the processor to a network environment; and in response to determining the data are in condition to be propagated beyond the processor: perform a data integrity check on the data; receive a result of a second data integrity check from a secondary computer system; compare the result of the data integrity check to the result of the second data integrity check; and, based at least in part on determining the result of the data integrity check matches the result of the second data integrity check, propagate the data beyond the processor. Corresponding systems and methods are also disclosed.