Dynamic Resource Control for Multi-Node Database Failover

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

In multi-node computer systems, when one node fails, the takeover node often lacks sufficient CPU resources to handle its own workload and the additional workload, leading to potential failures and increased failures across the system due to unaware application servers and user applications increasing requests.

Innovation Solution

A resource control mechanism dynamically monitors and manages resource usage by implementing policies such as rejecting new database connections, prioritizing processes, and batching messages to reduce high resource usage, transitioning between normal and safe modes to ensure efficient resource distribution and prevent node failure.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If the takeover node accepts additional workload from failed nodes, then system reliability is improved, but the takeover node may exceed its CPU resource capacity and fail

Engineering Contradiction:

Improvesystem reliabilityVSAvoidCPU resource capacity

Core Design Contradiction:

ReliabilityVSQuantity of substance

Solution Approach 1:

The patent implements dynamic resource allocation by continuously monitoring CPU usage metrics and adjusting process priorities in real-time. The resource control mechanism transitions between normal and safe modes based on current resource availability, allowing the system to adapt to changing workload conditions and prevent takeover node failure

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system changes operational parameters by monitoring CPU usage thresholds and transitioning resource allocation modes. When CPU usage exceeds safe thresholds, the system shifts from normal operation to safe mode, adjusting process priorities and resource distribution to maintain system reliability under resource constraints

Inventive Principle:

Principle #35Parameter changes

2Productivity

If application servers increase requests to maintain service levels, then service availability is improved, but more nodes may fail due to resource overload

Engineering Contradiction:

Improveservice availabilityVSAvoidnode stability

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The patent implements a feedback mechanism where the resource control mechanism continuously monitors CPU usage metrics and communicates resource status back to the workload distribution system. This feedback loop enables dynamic adjustment of request routing and throttling to prevent node overload while maintaining service availability

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The system performs preliminary resource assessment before accepting additional workload. The resource control mechanism evaluates current CPU usage and predicts resource requirements for incoming requests, preventing overload by rejecting or throttling requests that would push the node beyond safe operating limits

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS9128895B2Intelligent flood control management

Publication Date: 2015.09.08 ORACLE INT CORP

AI summary

Described herein are techniques for dynamically monitoring and managing resource usages of processes running on a node in a multi-node database system. High resource usages of processes can be proactively detected and alleviated, thereby making such a node to perform significantly better than otherwise.