Run clj-gatling load tests on your local machine or by utilizing AWS Lambda technology.
Note! This is an experimental alpha stage tool/library.
- Test your system using realistic scenarios. Not just one url.
- Write test scenarios in Clojure. No special DSL.
- Run your tests either from local machine or using multiple nodes using AWS Lambda technology.
Create new Clojure project & add the following to your project.clj
:dependencies
:
[clojider "0.5.1"]
Add this setting to your project.clj
:main clojider.core
Note! Clojider has to setup one S3 bucket, IAM role & policy and Lambda function using your AWS credentials. S3 bucket name you have to configure, other resources will be auto-named. The credentials are read from standard environment variables or configuration file. See details from here.
Add this setting to your project.clj
. This is required because report generation uses Gatling report generation module which is Scala code.
Report generation happens in local machine and this setting prevents Scala code to be included in to Jar file that is deployed to AWS Lambda environment.
:uberjar-exclusions [#"scala.*"]
Deploy your project to AWS Lambda. Note! Lambda is available in these regions: eu-west-1, us-east-1, us-west-2 and ap-northeast-1.
lein uberjar
lein run install -r <lambda-region> -b <s3-bucket-name> -f target/<your-uberjar-path>
You can find few simple examples here which you run locally in a following way.
lein run load-local -c 5 -d 10 -b <s3-bucket-name> -s clojider.examples/ping-simulation
or
lein run load-local -c 5 -d 10 -b <s3-bucket-name> -s clojider.examples/metrics-simulation
See clj-gatling on how to define test scenarios.
lein run load-local -c <concurrency> -d <duration-in-seconds> -s <simulation-symbol>
lein run load-lambda -r <lambda-region> -b <s3-bucket-name> -c <concurrency> -d <duration-in-seconds> -s <simulation-symbol>
And when you have updated your simulation (the scenario code), you have to update latest code to Lambda via
lein uberjar
lein run update -r <lambda-region> -b <s3-bucket-name> -f target/<your-uberjar-path>
-t
or `--timeout`` specifies request timeout in milliseconds. By default it is 5000 ms.
This will uninstall all created AWS resources (S3 bucket, role, policy and Lambda function). The cost of keeping these available in your account for the next load testing run is almost zero. Lambda pricing is totally based on the usage and S3 bucket contains only smallish old result files. However, I still wanted to have an option to destroy everything when you don't need the tool anymore.
lein run uninstall -r <lambda-region> -b <s3-bucket-name>
Use GitHub issues and Pull Requests.