Database Workload Throttle Adjustment via Rolling Average Metrics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current database management systems face challenges in dynamically adjusting throttles without requiring expensive state changes, leading to latency issues during workload management performance problems.
Innovation Solution
Implementing a method that monitors workload metrics and adjusts concurrency throttles based on theoretical and rolling average values, using control theory and statistical processing to dynamically adjust throttle levels without state changes, allowing for automatic query release and management of concurrency levels.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional throttle adjustment methods are used in database management systems, then workload management can be performed, but state changes are required which cause latency and performance issues
Solution Approach 1:
The patent implements dynamic throttle adjustment by continuously monitoring workload metrics (CPU usage, memory consumption, I/O operations) and automatically adjusting concurrency throttles in real-time without requiring state changes. The system uses a feedback loop that compares current metric values against target values and adjusts throttles accordingly, enabling proactive workload management that adapts to changing conditions without interruption.
Solution Approach 2:
The system monitors workload metrics and uses this feedback to automatically adjust throttle levels. By continuously measuring actual performance against target values and using this information to dynamically modify concurrency throttles, the system creates a closed-loop control mechanism that eliminates the need for manual state changes and reduces latency in workload management.
2Reliability
If manual throttle adjustment is performed, then workload management is achieved, but system response time increases due to state changes
Solution Approach 1:
The database management system performs self-adjustment by automatically monitoring its own workload metrics and adjusting throttle levels without external intervention. The system monitors CPU usage, memory consumption, and I/O operations, compares these against target values, and autonomously modifies concurrency throttles to maintain optimal performance, eliminating delays associated with manual state changes.
Solution Approach 2:
The system proactively adjusts throttles based on predicted workload conditions by continuously monitoring metrics and comparing them against target values before problems occur. This preliminary action allows the system to prevent performance degradation rather than react to it, maintaining faster response times and higher reliability.
3Device complexity
If simple monitoring is implemented, then system complexity is reduced, but the ability to dynamically adjust based on rolling average metrics is lost
Solution Approach 1:
The system dynamically adjusts throttle levels by incorporating rolling average metrics into its monitoring and control logic. Instead of using simple static thresholds, the system continuously updates rolling averages of workload metrics and uses these to dynamically modify concurrency throttles, enabling adaptable response to changing workload patterns while maintaining manageable system complexity through automated control.
Data Source
AI summary
A database system monitors a metric for a Workload Definition (WD) and adjusts a metric throttle for the WD to a new metric throttle level Cn, computed as the average of a theoretical metric level Cc that would drive a metric to a target T and a theoretical metric level Cr that would drive a rolling average of the metric to the target T.