Cloud Control Server Resource Scheduling for Service Availability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In conventional private cloud environments, resources are scheduled without considering the impact on services with the same service attribute, leading to downtime issues where multiple virtual machines with the same service attribute are scheduled on the same host computer, affecting overall service availability.
Innovation Solution
A resource scheduling method that involves a cloud control server and cluster control server to identify target host computers with no matching virtual machine labels, allowing virtual machines with the same service attribute to be deployed on different hosts, ensuring high availability by dispersing service attributes across multiple hosts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If virtual machines with the same service attribute are scheduled on the same host computer to improve resource utilization, then resource utilization is improved, but service availability deteriorates due to potential service disruptions during host downtime
Solution Approach 1:
The patent segments virtual machines by their service attributes and distributes them across different host computers. The scheduling system identifies the service attribute of each virtual machine and ensures that virtual machines with the same service attribute are placed on different hosts, creating a segmented distribution that prevents single points of failure while maintaining resource utilization.
Solution Approach 2:
The patent applies local quality by making the scheduling decision specific to each virtual machine's service attribute. Rather than applying a uniform scheduling rule, the system tailors the placement decision based on the local characteristic (service attribute) of each virtual machine, ensuring optimal distribution that balances resource utilization with service availability.
2Reliability
If virtual machines are dispersed across different host computers to ensure high availability, then service availability is improved, but resource utilization deteriorates due to potential idle capacity on hosts
Solution Approach 1:
The patent implements dynamic scheduling that adapts to the current state of host computers and virtual machine requirements. The system continuously evaluates resource availability and service attribute distributions, making flexible placement decisions that optimize both service availability and resource utilization based on real-time conditions rather than fixed rules.
Solution Approach 2:
The patent changes the scheduling parameter from simple resource-based allocation to service-attribute-aware placement. By introducing service attribute as a key scheduling parameter, the system transforms the allocation strategy to simultaneously achieve high availability through dispersion and high resource utilization through intelligent packing of compatible workloads.
3Device complexity
If a simple scheduling algorithm is used to reduce system complexity, then device complexity is reduced, but the ability to ensure high availability deteriorates
Solution Approach 1:
The patent extracts the service attribute characteristic from virtual machines and uses it as the primary scheduling criterion. By taking out this key attribute and making it the basis for scheduling decisions, the system achieves high availability guarantees without requiring complex algorithms, as the core logic focuses on matching service attributes to host assignments.
Solution Approach 2:
The patent performs preliminary classification of virtual machines by their service attributes before the actual scheduling placement. This preliminary action organizes the scheduling task in advance, allowing the system to make straightforward placement decisions that ensure high availability without needing complex real-time algorithms during the actual scheduling execution.
Data Source
AI summary
This application relates to a resource scheduling method, a resource scheduling system, a server, and a storage medium. The resource scheduling method includes receiving a virtual machine application request sent by a terminal, wherein the virtual machine application request includes a target virtual machine label. The resource scheduling method further includes comparing the target virtual machine label with a current virtual machine label of each host computer in a cluster to determine a target host computer, wherein the target host computer includes no virtual machine label matching the target virtual machine label, enabling the target host computer to create a first virtual machine, and setting a label of the first virtual machine as the target virtual machine label.