Fault-Tolerant Memory Copying via Modified Page Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional methods for replacing components in fault-tolerant computer systems require halting the entire system to copy memory, leading to inefficient downtime, which is unacceptable for real-time applications.
Innovation Solution
The system iteratively copies memory from one computer to another while allowing the first system to continue executing processes, tracking and copying only modified pages, and applying them to the second system, eventually allowing both to operate in lockstep with minimal downtime.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If traditional memory copying method is used, then memory can be copied to replacement system, but system must be halted causing unacceptable downtime
Solution Approach 1:
The patent segments the memory copying process into two phases: a brownout period where the replacement system is brought online and a blackout period where lockstep is achieved. During the brownout, the system copies modified pages iteratively while allowing normal operation to continue, eliminating the need to halt the entire system for memory copying.
Solution Approach 2:
The patent performs preliminary actions by tracking modified pages during a brownout period before the final blackout. The system identifies and copies only the modified pages that need to be transferred, rather than copying entire memory contents, thereby reducing the time required to prepare the replacement system.
2Productivity
If entire memory is copied during system halt, then replacement system can be brought online, but system operations are interrupted
Solution Approach 1:
The patent extracts only the modified pages from the entire memory content and copies just those to the replacement system. By taking out only the necessary changed portions rather than the entire memory, the system minimizes data transfer time and allows operations to continue during the brownout period.
Solution Approach 2:
The patent applies partial action by copying only the modified pages that are actually needed, rather than copying the entire memory contents. This partial copying approach reduces the time required to prepare the replacement system while maintaining sufficient accuracy for fault tolerance.
3Loss of time
If iterative memory copying is used, then downtime is minimized, but system complexity increases
Solution Approach 1:
The patent implements feedback by tracking which pages have been modified since the last copy operation. The system uses this feedback information to determine which pages need to be copied in subsequent iterations, reducing the amount of data transfer required over time and managing the complexity through intelligent selection.
Solution Approach 2:
The patent changes the parameter of what is being copied from entire memory contents to only modified pages. By changing this parameter dynamically based on which pages have changed, the system reduces copying time while managing complexity through targeted data transfer.
Data Source
AI summary
A method for tracking modified pages is provided. The method is utilized in a fault-tolerant system, the fault-tolerant system comprising a first computer system and a second computer system. The method involves copying a memory from the first computer system to the second computer system at a first point in time. Between the first point in time and a second point in time, a scheduler tracks processes that execute on the first computer system in a harvest list. After the second point in time, the processes that made changes to the memory between the first and second points in time are retrieved. The changes to the memory are then copied to the second computer system and the changes are applied to the memory of the second computer system.