Lifemapper Testing Tool (lmtest)

A tool for performing automated tests with Python.

This tool utilizes a test controller that schedules, runs, and evaluates tests. Additional documentation can be found at the lmtest GitHub pages.

Table of Contents

• Background
• Install
• Usage
  • Start Test Controller
  • Stop Test Controller
  • Creating Test Classes
  • Adding Tests to Run
• Contributing
• License

Background

This tool was created to monitor our servers to catch unexpected failures in our provided web services, database connections, machine conditions, etc. We find this tool to be useful for monitoring these elements and catching transient problems that may evolve over time (example: Full disk or changes in connected APIs).

Install

pip install lmtest-1.0.0-py3-none-any.whl

Usage

Start controller

run_controller -t TEST_DIRECTORY -l LOG_FILE start

This command starts the test controller using the tests configured in the TEST_DIRECTORY directory and logs notifications to LOG_FILE.

Stop controller

run_controller stop

This command will stop the controller.

Creating Test classes

Test classes inherit from the LmTest class found in lmtest.base.test_base.py. Subclasses should overwrite the run_test method to perform the desired test. If the test should add a new test to the schedule, use the add_new_test method and the test controller will add the resulting test(s) once the test is finished. Tests should raise lmtest.base.test_base.LmTestFailure if there is an error and lmtest.base.test_base.LmTestWarning if there is a warning. For both errors and warnings a meaningful message should be returned that will be emitted through the notification system.

Adding tests to run

The test controller utilizes test configuration files to determine which tests to run. These files should be located in the TEST_DIRECTORY referenced by the -t argument to the run_controller script. These test configuration files are in JSON format and have the required parameters, module, test_class, and parameters. The module parameter should be the package where the test class you wish to run is found. The test_class parameter is the test class in the specified module to run. Parameters should be a dictionary of parameters that will be used to instantiate the test class.

An example test configuration is:

{
    "module": "lmtest.tests.memory_usage_test",
    "test_class": "MemoryUsageTest",
    "parameters": {
       "warn_percent": 80,
       "error_percent": 95,
       "delay_time": 0,
       "delay_interval": 300
    }
}

Other example test configurations can be found in The example tests directory.

Contributing

We welcome contributions! See our current issues and please read our Contributing guide and our code of conduct.

License

See GPL3 License