Service Mesh Event Simulation for Cascading API Call Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Developers of extension applications in Software-as-a-Service (SaaS) systems face challenges in understanding the performance impact and costs of their extensions in production use, as existing testing methods are inadequate for evaluating cascading actions across multiple software systems and extension applications, which can lead to adverse performance and cost issues.
Innovation Solution
A simulation platform that processes event data and API monitoring data in combination with application process models and actual production data to generate simulation results, including call graphs and cardinality analysis, allowing for the evaluation of event-driven extensions and their cascading effects across multiple software systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If extension applications are deployed to provide additional functionality in SaaS systems, then system versatility and functionality are improved, but performance impact and costs become difficult to predict and control
Solution Approach 1:
The simulation platform performs preliminary evaluation of extension applications before deployment to production environments. It simulates the behavior of extension applications using monitored event data and API call patterns to predict performance impact and costs, allowing developers to identify and resolve issues before actual deployment.
Solution Approach 2:
The system implements a feedback loop by monitoring actual production data (events published and API calls received) and using this information to refine simulation models. This feedback mechanism improves the accuracy of performance predictions for future extension application evaluations.
2Productivity
If traditional testing methods are used for extension applications, then development time is reduced, but cascading actions across multiple software systems cannot be properly evaluated
Solution Approach 1:
The simulation platform acts as an intermediary between development testing and production deployment. It uses monitored production data as input to create realistic simulation scenarios that capture cascading actions across multiple software systems, providing accurate performance evaluation without requiring extensive production testing.
3Adaptability or versatility
If extension applications call APIs across multiple software systems, then system integration and functionality are enhanced, but cascading actions become hard to overlook and depend on data not accessible to developers
Solution Approach 1:
The simulation platform provides a universal evaluation environment that can simulate and analyze cascading actions across any number of software systems and extension applications. It consolidates data from multiple sources (event monitoring, API monitoring) into a comprehensive simulation model that makes hidden dependencies visible to developers.
Data Source
AI summary
Methods, systems, and computer-readable storage media for generating a call graph representative of a service mesh including software systems and extension applications, the software systems including a first software system that is configured to publish a first event and a second software system configured to receive first API calls, the extension applications including a first extension application configured to consume the first event and, in response to the first event, transmit first API calls, determining a first cardinality representative of a ratio of first events to first API-calls, receiving monitoring data representative of frequencies of the first event occurring during production use of the first software system, simulating production use of the service mesh by generating a set of frequencies of first API calls based on the call graph, the monitoring data, and the first cardinality, and displaying simulation results including the set of frequencies of the first API calls.