VM Workload Delay Control via Dirty Page Transfer Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Virtual machine fault tolerance mechanisms introduce additional delay in responding to user requests, which is problematic for delay-sensitive workloads like online games and real-time transactions, as the system response time cannot be guaranteed, leading to disconnection or transaction failure.
Innovation Solution
A data processing method that estimates the remaining processing time based on the transfer amount of dirty pages collected before a preset time point, allowing dynamic adjustment of the running operation's timing and determining the start time for subsequent operations, thereby controlling the delay time within a fixed range.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the virtual machine fault tolerance mechanism executes complete running operation, snapshot operation, and transfer operation before responding to user requests, then the fault tolerance reliability is improved, but the system response time increases causing delay-sensitive workload failures
Solution Approach 1:
The patent applies preliminary action by pre-calculating the remaining processing time based on the transfer amount of dirty pages collected before a preset time point. This allows the system to determine in advance whether the running operation can be completed within the required delay time, enabling proactive decision-making to meet response time requirements while maintaining fault tolerance reliability
Solution Approach 2:
The patent implements dynamics by dynamically adjusting the timing and duration of the running operation based on real-time calculations. The system suspends the running operation at a preset time point, calculates remaining processing time based on dirty page transfer amount, and continues or suspends the operation accordingly. This dynamic adjustment allows the system to adapt to varying workload conditions and meet delay-sensitive requirements while preserving fault tolerance capabilities
2Loss of time
If the running operation is suspended and continued multiple times to control delay time, then the system response time is improved, but the operation complexity increases
Solution Approach 1:
The patent applies feedback by continuously monitoring the transfer amount of dirty pages and using this information to calculate remaining processing time. This feedback mechanism enables the system to make informed decisions about when to suspend and continue the running operation, controlling delay time through data-driven adjustments rather than arbitrary complexity
Solution Approach 2:
The patent implements parameter changes by modifying the timing parameters of the running operation based on calculated remaining processing time. The system changes the duration and suspension points of operations dynamically, allowing control of delay time through parameter adjustment rather than fundamental structural changes, thereby managing complexity effectively
Data Source
AI summary
A data processing method, a data processing system and a computer program product are provided. The data processing method includes executing a running operation. The data processing method also includes suspending the running operation at a preset time point in a period of the running operation and calculating a remaining processing time according to a transfer amount of a plurality of dirty pages which are collected before the preset time point in the period of the running operation. The data processing method further includes continuing to execute the running operation, suspending the running operation, and executing a snapshot operation to generate a corresponding data snapshot based on the remaining processing time, and executing a transfer operation to transfer the corresponding data snapshot.