Locust.io Docker Image
This Docker image allows you to run the Locust.io load testing tool. It's different from other Locust-related Docker images in that it loads the actual test execution script (locustfile.py) from AWS S3. I.e., you can use the same Docker image for different load-testing scenarios without providing the locustfile.py "locally" on the server (or in the Docker image itself).
Starting a Locust Master
Set the LOCUST_MASTER environment variable to a non-empty value to start a master:
$ docker run -it --net=host -e LOCUST_MASTER=true -e LOCUST_FILE=s3://mybucket/mypath/locustfile.py registry.opensource.zalan.do/stups/locust:0.1
Starting a Locust Slave
Set the LOCUST_MASTER_HOST environment variable to the master's address (IP) to
start slave processes:
$ docker run -it --net=host -e LOCUST_MASTER_HOST=127.0.0.1 -e LOCUST_FILE=s3://mybucket/mypath/locustfile.py registry.opensource.zalan.do/stups/locust:0.1
Configuration Options
LOCUST_FILE- Filename of Locust test. The filename can reference a file on a S3 bucket. For example:
s3://my-bucket/my-folder/locustfile.py. LOCUST_TARGET_HOST- Host to load test. Example:
http://example.org LOCUST_MASTER- Run as Locust master
LOCUST_MASTER_HOST- Host of Locust master
LOCUST_CLIENTS- Number of users
LOCUST_HATCH_RATE- Number of new users per second
LOCUST_SLAVE_COUNT- Number of slaves to spawn (defaults to number of CPU cores)