Database Scheduling via Threshold Migration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing database flexible scheduling technologies face issues with high computational jitter, significant migration costs, and inefficient computation due to sensitivity to performance changes and pure mathematical optimization approaches, leading to instability and resource wastage in cloud computing environments.
Innovation Solution
A method and apparatus for database scheduling that involves obtaining performance parameters, determining if they match preset conditions, and migrating database instances to optimize workload distribution, reducing computational jitter and resource wastage by setting predetermined performance ranges and utilizing expectation factors for customized configurations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If optimization-based scheduling methods are used to achieve optimal distribution, then scheduling goal is improved, but computational jitter increases and stability deteriorates
Solution Approach 1:
The patent changes the scheduling approach from continuous optimization-based parameter adjustment to rule-based threshold triggering. When performance parameters (CPU usage, memory usage, I/O operations) exceed predetermined thresholds, scheduling actions are triggered. This parameter change eliminates continuous recomputation and associated jitter, improving service stability while maintaining scheduling effectiveness.
Solution Approach 2:
The patent segments the scheduling decision-making process into distinct performance metrics (CPU usage, memory usage, I/O operations) with separate thresholds. Each metric can independently trigger scheduling actions, allowing granular control over when migrations occur. This segmentation prevents unnecessary recomputations and reduces computational jitter compared to holistic optimization approaches.
2Productivity
If optimization-based scheduling methods are used to achieve optimal distribution, then scheduling goal is improved, but migration cost increases
Solution Approach 1:
The patent changes from continuous optimization that frequently triggers migrations to threshold-based triggering that only activates when performance degradation is detected. This reduces the frequency of migrations significantly, lowering the energy cost and operational overhead associated with instance migrations while still achieving scheduling goals when necessary.
Solution Approach 2:
The patent employs simple rule-based scheduling decisions rather than complex continuous optimization computations. The scheduling logic uses straightforward threshold comparisons instead of expensive mathematical optimization, making the scheduling mechanism itself computationally inexpensive and reducing overall system resource consumption.
3Productivity
If optimization-based scheduling methods are used to achieve optimal distribution, then scheduling goal is improved, but computation complexity increases
Solution Approach 1:
The patent extracts the essential scheduling decision-making from complex optimization algorithms and implements it through simple rule-based logic. By taking out only the necessary performance monitoring and threshold comparison functions, the system achieves scheduling without the computational burden of full optimization algorithms, reducing complexity while maintaining effectiveness.
Solution Approach 2:
The patent replaces expensive continuous optimization computations with inexpensive threshold-based decisions. Each scheduling decision requires only simple parameter comparisons rather than solving complex mathematical optimization problems, dramatically reducing computation complexity and making the system more scalable.
Data Source
AI summary
Apparatuses and methods are disclosed for database scheduling. An exemplary method may include obtaining a performance parameter associated with a database server. The method may also include determining whether the performance parameter matches a preset condition. After it is determined that the performance parameter matches the preset condition, the method may further include selecting a database instance operating on the database server and migrating the database instance to a receiving database server.