Cloud Host Restart Batching for VM Service Continuity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for restarting hosts in batches fail to efficiently manage service interruptions and ensure that the service is not interrupted.
Innovation Solution
The method involves obtaining a correspondence between virtual machines and hosts, and a maximum proportion of a quantity of virtual machines that can be shut down by network elements, and performing simulated live migration evaluation on conflicting virtual machines to minimize the duration required for service upgrades without interruptions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If all virtual machines are shut down before host restart, then the cloud OS can be upgraded to a later version, but the service interruption duration exceeds 30 minutes and service reliability is reduced
Solution Approach 1:
The patent divides the host restart process into multiple batches, where each batch contains a subset of virtual machines. By segmenting the restart operation, the system can maintain some virtual machines running while others are restarted, thus reducing overall service interruption duration and maintaining service reliability through parallel batch execution.
Solution Approach 2:
The patent performs preliminary classification of virtual machines into different batches before the restart process begins. This preliminary action allows the system to identify which virtual machines can be restarted together without affecting service continuity, enabling proactive planning that minimizes service interruption while ensuring reliable upgrades.
2Productivity
If virtual machines are restarted in large batches, then the upgrade process is faster, but the complexity of managing conflicting virtual machines increases
Solution Approach 1:
The patent performs preliminary classification of virtual machines into batches based on their compatibility and dependency relationships before the restart process. This advance organization simplifies the management of conflicting virtual machines during execution, allowing larger batches to be processed efficiently without increasing operational complexity.
Solution Approach 2:
The patent implements a feedback mechanism that evaluates the results of each batch restart and uses this information to optimize subsequent batch compositions. By analyzing the outcomes of previous batches, the system can adjust future batch sizes and compositions to maintain high upgrade speed while managing complexity effectively.
3Use of energy by moving object
If the maximum proportion of virtual machines is shut down, then resource utilization is optimized, but the risk of service disruption increases
Solution Approach 1:
The patent segments the virtual machine shutdown process into controlled batches, allowing the system to shut down the maximum proportion of virtual machines for resource optimization while maintaining service continuity through staged execution. Each batch is carefully managed to ensure service reliability is not compromised.
Solution Approach 2:
The patent applies partial action by shutting down virtual machines in controlled proportions rather than all at once or none at all. This approach allows the system to achieve resource utilization optimization by shutting down the maximum feasible proportion of virtual machines while maintaining service continuity through the staged, batched execution of shutdowns and restarts.
Data Source
AI summary
A method for restarting hosts in batches. Based on a virtual machine anti-affinity relationship, a correspondence between virtual machines and hosts is obtained. A maximum proportion of a quantity of virtual machines that can be shut down by network elements, a plurality of classification results between hosts and virtual machines in a to-be-restarted cloud OS, and conflicting virtual machines corresponding to each batch of hosts in each classification result is obtained. Simulated live migration evaluation on the conflicting virtual machines corresponding to each batch of hosts in each classification result of the plurality of classification results is performed to obtain a target classification result, where the target classification result is a classification result that has a smallest batch quantity in the plurality of classification results. A restart operation is performed on each batch of hosts in the cloud OS.