Application Service Dependency Detection via Pulse Timing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods fail to accurately determine application service dependencies in complex virtual data center environments, leading to difficulties in identifying failure points and efficiently recovering application services due to high load analysis costs and indeterminate results from changing application protocols.
Innovation Solution
A method and apparatus that acquire input and output pulses between application services, determine the time interval between these pulses, and establish application service dependencies based on whether a predetermined time interval is met, using a pulse fitting relationship to assess dependency, which allows for the identification of dependencies and the creation of a tree application service dependency chain.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If network traffic monitoring method is used to discover application service dependency, then dependency discovery is possible, but accuracy is insufficient in virtual data center environments
Solution Approach 1:
The patent introduces a dependency discovery agent as an intermediary component deployed on each virtual machine. This agent collects local service execution information and network traffic data, acting as a mediator between the virtualized application services and the central dependency discovery server. The agent enables accurate dependency tracking in virtual environments by providing structured local data that can be correlated with network traffic patterns, resolving the accuracy issue in virtual data centers.
Solution Approach 2:
The patent replaces traditional network traffic monitoring methods with a hybrid approach that combines automated agent-based data collection with machine learning algorithms. Instead of relying solely on passive network monitoring, the system uses agents to actively collect service execution metrics and combines this with ML-based analysis to automatically discover dependencies, improving both accuracy and adaptability to dynamic virtual environments.
2Reliability
If comprehensive load analysis is performed on network application services, then analysis completeness is improved, but analysis cost increases significantly
Solution Approach 1:
The patent implements a two-stage dependency discovery approach: first using lightweight rule-based methods for initial dependency identification, then applying machine learning algorithms only to ambiguous or complex cases. This partial application of computationally expensive ML techniques maintains analysis completeness while significantly reducing overall analysis cost compared to applying ML uniformly to all service interactions.
Solution Approach 2:
The patent segments the dependency discovery process into multiple stages: data collection by local agents, preliminary analysis using simple rules, and refined analysis using machine learning only when needed. This segmentation allows the system to achieve comprehensive analysis through a combination of lightweight and heavy-weight methods, optimizing the balance between completeness and computational cost.
3Ease of manufacture
If traditional dependency discovery methods are used, then implementation simplicity is maintained, but inability to meet changing application protocol requirements occurs
Solution Approach 1:
The patent designs the dependency discovery agent with universal functionality that can handle multiple application protocols and service types through a unified interface. The agent collects standardized service execution information regardless of the specific protocol being used, and the machine learning model learns protocol-agnostic dependency patterns. This universal design maintains implementation simplicity while providing adaptability to changing protocol requirements.
Solution Approach 2:
The patent implements a dynamic dependency discovery system where the machine learning model continuously learns from new service interactions and adapts to changing protocols. The system evolves its understanding of dependencies over time, allowing it to maintain simplicity in implementation while becoming increasingly adaptable to new application protocols and service patterns without requiring manual reconfiguration.
Data Source
AI summary
The present disclosure provides a method and apparatus for determining an application service dependency, and a processor. The method includes that: an input pulse of first application service and an output pulse of second application service are acquired; a time interval between the input pulse and the output pulse is determined; and an application service dependency between the first application service and the second application service is determined based on the time interval. According to the present disclosure, a technical problem that an application service dependency between different pieces of application service cannot be determined accurately through a method for determining an application service dependency provided in the related art is solved.