> For the complete documentation index, see [llms.txt](https://rajadilipkolli.gitbook.io/my-spring-boot-experiments/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://rajadilipkolli.gitbook.io/my-spring-boot-experiments/boot-chaos-monkey.md).

# spring-boot-chaos-monkey

[Choas Monkey](https://netflix.github.io/chaosmonkey/)

This module explores Chaos Engineering principles by intentionally injecting failures into a Spring Boot application to test its resilience. It demonstrates how to set up and integrate `chaos-monkey-spring-boot` for your microservices, as well as performance testing with Gatling to measure any degradation under chaos scenarios.

![](/files/pu7D3sHfTSCCe6r9QsTw)

***

## Key Features

1. **Chaos Injection**: Simulate random latencies, exceptions, or resource unavailability.
2. **Performance Testing** (Gatling): Benchmark throughput and latency under chaotic conditions.
3. **Monitoring**: Validate that chaos is functioning by observing logs, metrics, or custom dashboards.

***

### Run tests

```shell
./mvnw clean verify
```

### Run Gatling

Bring the application up and then start Gatling

```shell
./mvnw gatling:test
```

### Run locally

```shell
docker-compose -f docker/docker-compose.yml up -d
./mvnw spring-boot:run -Dspring-boot.run.profiles=local -Dspring-boot.run.arguments="--chaos.monkey.enabled=true"
./mvnw gatling:test
```

**Verification**:

* Check application logs for entries containing "Chaos Monkey" to confirm injection events
* Use Gatling or other load testing tools to measure:
  * Response time degradation
  * Error rate changes
  * System resource utilization

By default, Chaos Monkey is enabled. Let's disable it and run the tests again. It should improve the response times.

```shell
./mvnw spring-boot:run -Dspring-boot.run.profiles=local -Dspring-boot.run.arguments="--chaos.monkey.enabled=false"
./mvnw gatling:test
```

Results of gatling are stored in target folder for each run

### Useful Links

* Swagger UI: <http://localhost:8080/swagger-ui.html>
* Actuator Endpoint: <http://localhost:8080/actuator>
* PGAdmin : <http://localhost:5050>
* Obervability metrics : <http://localhost:8080/actuator/metrics>

### Technologies used

* Spring Boot
* Gatling (Performance Tests)
* Chaos Monkey (Chaos Engineering Principles)

#### Sample Test Results

We can observe that mean requests/sec is more when Chaos Monkey is enabled.

### With Chaos Monkey Enabled

\================================================================================ ---- Global Information --------------------------------------------------------

> request count 6000 (OK=6000 KO=0 ) min response time 1 (OK=1 KO=- ) max response time 2635 (OK=2635 KO=- ) mean response time 145 (OK=145 KO=- ) std deviation 337 (OK=337 KO=- ) response time 50th percentile 15 (OK=15 KO=- ) response time 75th percentile 98 (OK=98 KO=- ) response time 95th percentile 773 (OK=773 KO=- ) response time 99th percentile 1952 (OK=1952 KO=- ) mean requests/sec 193.548 (OK=193.548 KO=- ) ---- Response Time Distribution ------------------------------------------------ t < 800 ms 5733 ( 96%) 800 ms <= t < 1200 ms 131 ( 2%) t ≥ 1200 ms 136 ( 2%) failed 0 ( 0%) ================================================================================

**With Chaos Monkey disabled**

\================================================================================ ---- Global Information --------------------------------------------------------

> request count 6000 (OK=6000 KO=0 ) min response time 4 (OK=4 KO=- ) max response time 13608 (OK=13608 KO=- ) mean response time 6289 (OK=6289 KO=- ) std deviation 2579 (OK=2579 KO=- ) response time 50th percentile 6396 (OK=6396 KO=- ) response time 75th percentile 8292 (OK=8292 KO=- ) response time 95th percentile 10342 (OK=10342 KO=- ) response time 99th percentile 11604 (OK=11604 KO=- ) mean requests/sec 176.471 (OK=176.471 KO=- ) ---- Response Time Distribution ------------------------------------------------ t < 800 ms 65 ( 1%) 800 ms <= t < 1200 ms 54 ( 1%) t ≥ 1200 ms 5881 ( 98%) failed 0 ( 0%) ================================================================================