The RAFT framework provides a flexible environment for engineering-level device testing, empowering developers to thoroughly validate code changes before commits, promoting code quality and enabling the testing of multiple commits simultaneously.
Contents
The system includes a flexible parsing mechanism for configuration. This allows simple environment configuration for describing racks, the devices within them, network settings and other essential test parameters. Providing ultimate flexibility in testing setup, without needing to modify the test.
Test runs can be customised by providing dynamic command-line arguments. These arguments might specify which rack to use, device slots to target, debug mode, and other runtime settings. The system intelligently combines these arguments with your pre-defined configurations.
Centralized Device Management: The core of the system is a device manager that provides a centralized way to organize and control multiple devices in the testing setup.
It understands how to:
- Model Devices: Representing each device as an object, encapsulating its configuration details.
- Establish Consoles: Establishing different types of communication channels with devices, supporting protocols like SSH, Telnet, or Serial for management and control.
- Manage Power: Integrates with power control mechanisms to cycle the power state of devices as needed during testing.
- Handle Inbound/Outbound Connections: Configures devices to initiate inbound/outbound connections if required by your test scenarios.
- Remote Control: Provides a unified interface for sending remote control commands to your devices, abstracting differences between remote types.
- Minimal Infrastructure: Adaptive minimal testing infrastructure requiring no specialised hardware.
- Quick setup: With python and some pip packages installed, testing can begin.
- Flexibility: Supports various devices, communication protocols, and integrates with different power control systems.
- Ease of Use: Provides a clear interface to interact with devices, simplifying your test scripts and focusing on the test logic itself.
- Organization: Introduces structure to your test environment, making it easier to manage test cases and scenarios.
-
Python (=3.11.8)
- Pyenv can be used to manage multiple versions of Python (please refer to the documentation).
-
All the packages listed in requirements.txt
- Can be installed using
$ pip install -r requirements.txt
- Can be installed using
Clone the repository and ensure that all requirements are met.
For our first test script we will be using the example_test.py. There are extensive comments in the test files to explain what it does.
This test script can be run with the below command:
./example_test.py --config ../configs/getting_started_rack_config.yml
Two config files are used for running this test. They are:
- The Rack config
- This yaml file is used to define the setup your DUT (Device Under Test). In this file the connections, IP addresses and controllers connected to the device are listed.
- For first test we used the getting_started_rack_config.yml from the docs directory.
- For further information on the rack config see the example_rack_config.yml
- The device config
- This yaml file is used to define device types. The information defined in this config is consistent across all devices of a type.
- For of our first test we will use the example_device_config.yml
After running your test you should find a new folder has been created called logs. This should contain test logs in both text and .csv format.
The summary files in this directory show the results of all tests that ran.
If multiple instances of our test class had been created they would each get a test-<index>.log and corresponding .csv, with all their results being collated into the summary files.
See contributing file: CONTRIBUTING.md
See license file: LICENSE
<style> #toc { font-size: 1.5em; font-weight: 500; } </style>