Cloud VM Resource Sharing via Emergency Policy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing cloud computing systems face low utilization rates of virtual machine resources due to the pre-creation of standby virtual machines, leading to inefficient resource allocation and increased service fluctuations during deployment.
Innovation Solution
A cloud application processing method that introduces an emergency policy allowing a second virtual machine to temporarily lend resources to a first virtual machine, creating a new virtual machine only when necessary, thereby improving resource utilization and reducing service fluctuations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If standby virtual machines are pre-created for each deployed application, then service reliability is improved through immediate failover capability, but virtual machine resource utilization rate deteriorates due to idle standby resources
Solution Approach 1:
The patent applies universality by enabling virtual machines to serve multiple purposes: running their own applications and simultaneously serving as standby resources for other applications. The resource pool virtual machines can be dynamically allocated to provide backup capacity across multiple applications, making the standby infrastructure universal rather than application-specific.
Solution Approach 2:
The patent merges the standby virtual machine resources into a shared resource pool that serves multiple applications collectively. Instead of dedicating separate standby VMs to each application, the system combines standby resources into a unified pool that can be dynamically allocated to any application needing backup capacity, improving overall resource utilization while maintaining reliability.
2Speed
If standby virtual machines are pre-created, then failover speed is improved, but resource allocation efficiency deteriorates due to static resource reservation
Solution Approach 1:
The patent implements dynamics by transitioning from static pre-created standby VMs to dynamic on-demand standby resource allocation. The system continuously monitors resource availability and application needs, dynamically allocating standby capacity from the resource pool when required and releasing it when not needed, enabling both fast failover and efficient resource utilization.
Solution Approach 2:
The system enables self-service through automated standby resource allocation that responds to application needs without manual intervention. The resource pool automatically provides standby capacity to applications that require it, and the system self-manages the allocation and release of these resources based on real-time conditions, improving both response speed and allocation efficiency.
3Reliability
If new virtual machines are created for standby capacity, then service availability is improved, but service fluctuation deteriorates during the VM creation process
Solution Approach 1:
The patent applies preliminary action by pre-establishing the resource pool infrastructure and maintaining a baseline level of standby virtual machines before failures occur. This preliminary preparation ensures that backup capacity is already available or can be rapidly activated, avoiding service fluctuations that would occur during ad-hoc VM creation when failures happen.
Data Source
AI summary
A cloud application processing method and related apparatus are provided. The method is performed by a cloud service provider, and may include determining that a working state of a first virtual machine satisfies a condition for adding a virtual machine, determining, according to an emergency policy corresponding to a first application running on the first virtual machine, a second application that has an emergency relationship with the first application, and instructing a second virtual machine on which the second application is hosted to run the first application deployed on the second virtual machine, creating a third virtual machine, deploying and starting the first application on the third virtual machine, and instructing the second virtual machine to stop running the first application after the first application is started on the third virtual machine.