Application Instance Health Detection via Resource Usage Monitoring

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing methods fail to detect application instances operating improperly, particularly those that do not utilize computing resources correctly, leading to inefficient processing and potential 'storm drain' scenarios where malfunctioning instances handle a larger workload, undetected by external indicators.

Innovation Solution

A method is implemented to determine if an application instance has properly used computing resources by monitoring characteristics of workload requests, enabling policies to assess resource usage and initiate actions, such as disabling the instance, to prevent improper processing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Difficulty of detecting and measuring

If external indicators are used to detect malfunctioning application instances, then detection simplicity is maintained, but detection precision deteriorates because external indicators cannot detect instances that do not exhibit observable abnormalities

Engineering Contradiction:

Improvedetection simplicityVSAvoiddetection precision

Core Design Contradiction:

Difficulty of detecting and measuringVSMeasurement precision

Solution Approach 1:

The patent introduces computing resource usage data as an intermediary indicator to detect application instance health. Instead of relying solely on external indicators, the system monitors resource usage metrics (CPU, memory, I/O) as intermediate signals that reflect the internal state of application instances. This intermediary measurement mechanism enables detection of malfunctions that do not manifest externally.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent replaces the mechanical/external observation method with an information-based monitoring system. Instead of observing external behaviors or symptoms, the system substitutes this with direct monitoring of computing resource usage data through software agents and collectors, transitioning from physical observation to digital metric analysis.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If computing resource usage monitoring is implemented to detect improper operation, then detection precision improves, but device complexity increases due to additional monitoring infrastructure

Engineering Contradiction:

Improvedetection precisionVSAvoidsystem complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent implements self-service monitoring where application instances themselves report their computing resource usage to the workload manager. Each application instance acts as its own monitoring subject, collecting and reporting its own resource consumption data. This eliminates the need for external monitoring agents on each instance, reducing system complexity while maintaining detection precision.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The workload manager performs multiple functions: it manages workload distribution, tracks computing resource usage, detects malfunctioning instances, and initiates corrective actions. By consolidating these functions in a single system component, the patent avoids the complexity of separate specialized systems for each function, achieving multi-functionality that reduces overall system complexity.

Inventive Principle:

Principle #6Universality (Multi-functionality)

3Reliability

If malfunctioning application instances are allowed to continue processing, then service availability is maintained, but productivity deteriorates due to storm drain scenarios where faulty instances handle excessive workload

Engineering Contradiction:

Improveservice availabilityVSAvoidprocessing efficiency

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The patent implements a feedback loop where the workload manager continuously monitors computing resource usage data from application instances and uses this information to adjust workload distribution. When malfunctions are detected through resource usage anomalies, the system provides feedback by redirecting workloads away from affected instances, dynamically adapting to maintain both availability and efficiency.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The patent introduces dynamic workload management where the distribution of workloads is not static but continuously adjusted based on real-time computing resource usage monitoring. The system can dynamically identify and isolate malfunctioning instances by analyzing resource usage patterns, enabling adaptive response that maintains service availability while preventing productivity loss from storm drain scenarios.

Inventive Principle:

Principle #15Dynamics

Data Source

PatentUS10540210B2Detecting application instances that are operating improperly

Publication Date: 2020.01.21 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

A method includes determining whether an application instance has properly used computing resources to process a workload request. The determination is based on one or more characteristics of the workload request to be processed. The method can include determining whether the application instance is operating properly, based on the determining of whether the application instance has properly used computing resources. Based on the determination that the application instance is not operating properly, the method initiates an action to prevent the application instance from improperly processing workload requests.