Microsoft patterns & practices
This reference implementation shows a set of best practices for building and running a microservices architecture on Microsoft Azure, using Kubernetes.
This project has a companion set of articles that describe challenges, design patterns, and best practices for building microservices architecture. You can find these articles on the Azure Architecture Center:
Fabrikam, Inc. (a fictional company) is starting a drone delivery service. The company manages a fleet of drone aircraft. Businesses register with the service, and users can request a drone to pick up goods for delivery. When a customer schedules a pickup, a backend system assigns a drone and notifies the user with an estimated delivery time. While the delivery is in progress, the customer can track the location of the drone, with a continuously updated ETA.
The Drone Delivery application is a sample application that consists of several microservices. Because it's a sample, the functionality is simulated, but the APIs and microservices interactions are intended to reflect real-world design patterns.
- Ingestion service. Receives client requests and buffers them.
- Scheduler service. Dispatches client requests and manages the delivery workflow.
- Supervisor service. Monitors the workflow for failures and applies compensating transactions.
- Account service. Manages user accounts.
- Third-party Transportation service. Manages third-party transportation options.
- Drone service. Schedules drones and monitors drones in flight.
- Package service. Manages packages.
- Delivery service. Manages deliveries that are scheduled or in-transit.
- Delivery History service. Stores the history of completed deliveries.
The Drone Delivery application has been tested up to 2000 messages/sec:
| Replicas | ~Max CPU (mc) | ~Max Mem (MB) | Avg. Throughput* | Max. Throughput* | Avg (ms) | 50th (ms) | 95th (ms) | 99th (ms) | |
|---|---|---|---|---|---|---|---|---|---|
| Nginx | 1 | N/A | N/A | serve: 1595 reqs/sec | serve: 1923 reqs/sec | N/A | N/A | N/A | N/A |
| Ingestion | 10 | 474 | 488 | ingest: 1275 msgs/sec | ingest: 1710 msgs/sec | 251 | 50.1 | 1560 | 2540 |
| Workflow (receive messages) | 35 | 1445 | 79 | egress: 1275 msgs/sec | egress: 1710 msgs/sec | 81.5 | 0 | 25.7 | 121 |
| Workflow (call backend services + mark message as complete) | 35 | 1445 | 79 | complete: 1100 msgs/sec | complete: 1322 msgs/sec | 561.8 | 447 | 1350 | 2540 |
| Package | 50 | 213 | 78 | N/A | N/A | 67.5 | 53.9 | 165 | 306 |
| Delivery | 50 | 328 | 334 | N/A | N/A | 93.8 | 82.4 | 200 | 304 |
| Dronescheduler | 50 | 402 | 301 | N/A | N/A | 85.9 | 72.6 | 203 | 308 |
*sources:
- Serve: Visual Studio Load Test Throughout Request/Sec
- Ingest: Azure Service Bus metrics Incoming Messages/Sec
- Egress: Azure Service Bus metrics Outgoing Messages/Sec
- Complete: AI Service Bus Complete dependencies
- Avg/50th/95th/99th: AI dependencies
- CPU/Mem: Azure Monitor for Containers
To deploy the solution, follow the steps listed here.
This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact opencode@microsoft.com with any additional questions or comments.