DefinePK

Research Objective

To evaluate the performance of microservices architecture implemented in .NET Core, focusing on scalability, latency, and resource utilization compared to monolithic systems.

Methodology

A systematic approach evaluating microservices implemented in .NET Core, utilizing Docker for containerization, Kubernetes for orchestration, gRPC and RabbitMQ for communication, and Ocelot as an API Gateway. Testing involved simulating user traffic using Apache JMeter to measure scalability (throughput, error rate), latency under various loads, resource utilization (CPU, memory), and fault tolerance (recovery time). Data consistency was managed using the Saga pattern.

Methodology Flowchart

                        graph TD; A[Design & Setup: .NET Core Microservices, Docker, Kubernetes] --> B[Communication Setup: gRPC, RabbitMQ, Ocelot Gateway]; B --> C[Testing Scenarios]; C --> D[Scalability TestingJMeter]; C --> E[Latency Testing]; C --> F[Resource Utilization Monitoring]; C --> G[Fault Tolerance TestingSimulated Failures]; D --> H[Record Throughput/Errors]; E --> I[Record Response Times]; F --> J[Record CPU/Memory/Bandwidth]; G --> K[Record Recovery Time/Availability]; H & I & J & K --> L[Data Analysis & Comparison Tables]; L --> M[Conclusion on Performance Metrics];                    

Discussion

The study confirms that microservices in .NET Core offer superior scalability, faster response times, and better resource utilization for high-performance web applications by allowing independent scaling of services and utilizing efficient communication protocols (gRPC, RabbitMQ). The modularity minimizes resource contention. However, challenges remain regarding data consistency (requiring patterns like Saga) and the complexity of monitoring and security in distributed environments.

Key Findings

The microservices architecture demonstrated significant performance improvements: over double the throughput (8,400 req/s vs. 3,200 req/s) and concurrent user support (10,000 vs. 4,500) compared to the monolithic system. Latency was nearly halved under peak load (250 ms vs. 450 ms). Resource efficiency was enhanced, with lower average CPU usage (65% vs. 85%) and memory usage. Fault tolerance was superior, showing faster average recovery time (12 s vs. 35 s) and higher availability (99.9% vs. 96.5%).

Conclusion

Microservices architecture implemented with .NET Core provides a robust, scalable, and efficient solution for dynamic, high-demand web environments, significantly outperforming monolithic systems across key performance indicators, particularly scalability and fault tolerance.

Fact Check

1. Maximum Concurrent Users supported by Microservices Architecture: 10,000. (Confirmed in Table 1)
2. Average Throughput for Monolithic Architecture: 3,200 requests/s. (Confirmed in Table 1)
3. Average Recovery Time for Microservices Architecture: 12 seconds. (Confirmed in Table 4)

Mind Map