There are a few different ways to configure flood-agent, though you definitely don't need to know them all in detail to start running load tests. These options are provided to make life easier for advanced users and operations teams.

To get started, use the guided flood-agent configure configuration wizard, or drop a file called config.yaml file into the same directory as flood-agent.

If you need to take advantage of the more advanced configuration options, read on below for details.

Configuration basics

Agent configuration has three sources, applied in order of priority

1. command line flags

2. environment variables

3. the configuration file

Flags take highest precedence, followed by environment variables and then the config file.

These different sources may each be useful in different automated configuration management approaches.

For example, you may choose to produce a base static configuration file for all agent installations, and override values on specific hosts using environment variables.

Tool config

Tool specific config is more detailed, and so can only be specified in the config file.

If the tool configuration is left blank, and docker is available, the default is to use docker containers for any of available the load generators (JMeter, Gatling, Flood Element and Selenium).

However, only JMeter and Gatling are currently able to be run directly.

Tools are not automatically detected. To generate load using a particular tool, you must first set it up in the config file.

For details please see the tool-specific page:

• ​JMeter​

• ​Gatling​

Config file search paths

To facilitate package management and configuration management tools, the configuration file is searched for in a number of standard locations:

1. ./config.yaml (the current directory)

2. $HOME/.config/flood-grid/config.yaml

3. /etc/flood-grid/config/config.yaml