Database Regulator Adjusting Request Priority via Optimizer Estimates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional database management systems fail to effectively respond to various internal or external events impacting performance and do not properly consider the impact of inaccurate optimizer estimates on meeting service level goals for diverse workloads.
Innovation Solution
A database system with an optimizer that generates resource estimates and a regulator subsystem that dynamically adjusts priority settings and resource allocation based on these estimates to achieve service level goals, utilizing a closed-loop workload management architecture with both macro- and micro-regulators to manage and adjust database resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional database management systems use fixed priority scheduling, then system simplicity is maintained, but service level goals cannot be consistently met under varying workloads
Solution Approach 1:
The patent implements dynamic priority adjustment where the regulator continuously monitors workload characteristics and optimizer estimates, then adjusts request priorities in real-time based on current system conditions. This transforms the static priority scheduling into a dynamic system that adapts to varying workloads, ensuring service level goals are met while managing complexity through automated feedback loops
Solution Approach 2:
The regulator subsystem establishes a closed-loop feedback mechanism that continuously monitors workload execution, compares actual performance against service level goals, and adjusts priorities accordingly. The system uses feedback from optimizer resource estimates and actual workload performance to dynamically regulate priority levels, enabling reliable service level achievement without manual intervention
2Reliability
If the system dynamically adjusts priority levels based on optimizer estimates, then service level goal achievement improves, but system complexity increases
Solution Approach 1:
The regulator subsystem acts as an intermediary layer between the optimizer and the workload execution system. It receives resource estimates from the optimizer, processes this information through regulation logic, and adjusts priorities accordingly. This intermediary structure isolates the complexity of dynamic regulation from both the optimizer and the core execution engine, allowing service level goal improvement without proportionally increasing overall system complexity
3Productivity
If conventional systems use static resource allocation, then system simplicity is maintained, but productivity decreases under diverse workloads
Solution Approach 1:
The system transitions from static resource allocation to dynamic resource management where the regulator continuously adjusts priority levels based on real-time workload characteristics and optimizer estimates. This dynamic allocation optimizes productivity for diverse workload types (batch, interactive, background) by automatically distributing resources according to current system conditions and service level requirements
Solution Approach 2:
The regulator changes the priority parameter of workloads dynamically based on monitored execution progress and optimizer resource estimates. By adjusting this key parameter in response to system conditions, the system optimizes productivity across different workload types without requiring complex manual resource management, achieving efficient resource utilization through automated parameter adjustment
Data Source
AI summary
A database system includes an optimizer to generate resource estimates regarding execution of a request in the database system, and a regulator to monitor execution of a request and to adjust a priority level of the request based on the monitored execution and based on the resource estimates provided by the optimizer.