Cloud Server Scheduling via Time-Based Cluster Migration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current cloud server deployment methods lead to resource competition and inefficient utilization due to varying access patterns across different systems, resulting in unstable operations and resource wastage.
Innovation Solution
A method and apparatus for scheduling cloud servers by monitoring pre-set time periods to dynamically allocate idle servers from one cluster to another, enabling efficient task execution and resource utilization through backup and configuration management, including IP forwarding for network communication.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If different systems are deployed to the same cloud server cluster, then deployment cost is reduced, but resource competition occurs and system stability deteriorates
Solution Approach 1:
The patent implements dynamic resource allocation by allowing cloud servers to be scheduled between different clusters based on real-time task requirements and resource availability. The system dynamically adjusts which cluster a server belongs to depending on current workload patterns, resolving the contradiction between cost efficiency and stability by making the assignment flexible rather than fixed.
Solution Approach 2:
The patent changes the parameter of cluster assignment from static to dynamic based on time period and resource utilization patterns. By monitoring system access patterns and resource usage, the system adjusts cluster assignments to optimize both cost and stability, allocating resources differently at peak versus off-peak times.
2Reliability
If different systems are deployed to different cloud server clusters, then system stability is improved, but resource utilization efficiency deteriorates during off-peak periods
Solution Approach 1:
The system dynamically reallocates cloud servers between clusters based on monitored access patterns. During off-peak periods when certain systems have low traffic, the system can move those servers to different clusters to be utilized by other systems, preventing resource waste while maintaining stability during peak periods through the same dynamic mechanism.
Solution Approach 2:
The patent implements periodic monitoring and reallocation of cloud servers based on time periods. The system monitors access patterns at different times and adjusts server assignments accordingly, allowing resources to be optimized during off-peak periods while maintaining stable configurations during peak periods, thus reducing overall resource waste.
3Reliability
If cloud server resources are allocated based on maximum access amount, then sufficient resources are provided during peak periods, but resources are wasted during off-peak periods
Solution Approach 1:
The system dynamically adjusts resource allocation based on real-time monitoring of access patterns rather than relying on maximum capacity estimates. During peak periods, resources are allocated to ensure sufficient service capability, while during off-peak periods, the same dynamically allocated resources can be reassigned to other systems or put into idle state, eliminating waste while maintaining peak capability when needed.
Data Source
AI summary
The present disclosure provides a method and apparatus for scheduling a cloud server. A specific implementation mode of the method comprises: monitoring whether current time is in a first pre-set time period; in response to the monitoring that the current time is in the first pre-set time period, scheduling a cloud server in a first cloud server cluster having a running state being an idle state, as a target cloud server, to a second cloud server cluster, so that the target cloud server executes a task obtained by the second cloud server cluster; monitoring whether the current time is in a second pre-set time period; in response to the monitoring that the current time is in the second pre-set time period, rescheduling the target cloud server to the first cloud server cluster, so that the target cloud server executes a task obtained by the first cloud server cluster.