VM Migration Error Origin Identification via Distinct Memory Identifiers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In virtual machine migration, uncorrectable errors (UEs) detected during data transfer between migration source and destination lead to data corruption and abnormal termination, as it is difficult to distinguish between errors originating from the source and destination devices, causing operational issues.
Innovation Solution
A system where the migration source device notifies the migration destination of error information, allowing the destination to write a distinct identifier to memory addresses associated with erroneous data, enabling differentiation between source and destination errors, thus preventing data corruption and allowing continued operation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If error detection is performed during virtual machine migration, then data integrity can be monitored, but it becomes difficult to distinguish between source and destination errors, leading to migration termination
Solution Approach 1:
The patent applies preliminary action by writing a second identifier to the memory of the destination device before actual data migration occurs. This pre-positioned identifier enables immediate differentiation between source and destination errors once they are detected during migration, resolving the complexity of error identification while maintaining data integrity monitoring.
Solution Approach 2:
The patent introduces an intermediary mechanism in the form of a second identifier that acts as a marker in the destination device's memory. This intermediary element facilitates the distinction between errors originating from the source device versus those from the destination device, solving the problem of error source identification without requiring complex analysis during migration.
2Productivity
If migration continues despite detected errors, then operational continuity is maintained, but data corruption may occur without proper error differentiation
Solution Approach 1:
By pre-writing the second identifier to the destination memory before migration, the system enables continuous operation even when errors are detected. The identifier allows the system to distinguish between correctable and incorrectable errors, permitting migration to continue when appropriate while preventing data corruption through informed decision-making.
Solution Approach 2:
The patent implements feedback by using the second identifier to provide information about error origins during migration. This feedback mechanism enables the system to make informed decisions about whether to continue or terminate migration based on the specific error source, balancing productivity maintenance with data protection.
3Ease of operation
If traditional error handling is used without error source differentiation, then migration can be simplified, but operational stability is compromised due to inability to address specific error types
Solution Approach 1:
The pre-writing of the second identifier simplifies the migration process by automatically providing error source differentiation without requiring complex procedures. This preliminary action maintains operational simplicity while significantly improving reliability by enabling specific error handling based on the identified error source.
Solution Approach 2:
The second identifier serves as an intermediary that bridges the gap between simple error detection and complex error analysis. It provides automatic error source differentiation that maintains ease of operation while ensuring operational stability through targeted error handling procedures.
Data Source
AI summary
A system for migrating a virtual machine includes: a first device; and a second device. The first device notifies the second device of failure information indicating that data transported to the second device includes an error originated from the first device, when the error is detected from a storage area in the first device. The second device writes a second identifier, whose value is different from that of a first identifier, into a memory of the second device while associating the second identifier with a second address, the second address being to serve as a copy destination of data indicated by the failure information, wherein the first identifier indicates that data in the second address includes an error originated from the second device, and the second identifier serves as an identifier indicating that data in the second address includes an error originated from the first device.