Running Tests¶

Basic Operation¶

Within a test module with a etr.yaml configuration file etr needs no arguments and can be executed like this:

$ etr

Note

Use $ etr --help to get details on command line options.

To run with more information, pass the -v flag, which also can be passed more times for even more verbose output like this:

$ etr -v
$ etr -vv

Note

Verbosity level 2 (-vv) provides many details that are targeting a developer of etr and some messages may be confusing. It is recommended for normal users to use verbosity level 1 (-v) for normal usage.

To execute a test in another directory use the -C flag:

$ etr -C path/to/test/directory

Steps¶

etr execution is separated into three steps:

setup - Aquisition of test resources. - Process jinja2 template files.
run - Run tests.
teardown - Release test resources.

By default all steps are executed but can also be controlled via the --step option:

# Execute only the setup step
$ etr --step setup

# Execute only the setup and run steps to leave everything running so that run may be
# executed again
$ etr --step setup:run

# Execute tests again
$ etr -step run

Test Subsets¶

You can specify which tests from etr.yaml to run with the test option:

$ etr --test <test>
$ etr -t <test>

for example:

$ etr -t src/tests.robot

The option may be specified multiple times:

$ etr -t src/tests1.robot -t src/tests2.robot

Cleanup Logs¶

Run etr clean to cleanup logs, session information and more.

Caution

Take care to release acquired resources before doing this. Otherwise the locks will be lost and manual recovery will have to be made.

Resources are normally automatically released during during teardown. If that step was excluded from execution this step can be manually executed with etr --step teardown.

Additional Information¶

Orphaned Processes¶

Whenever etr runs a test with e.g. Robot Framework it will set the environment variable ETR_TRACE_ID with a unique identifier. This identifier is normally inherited if further processes are started from Robot tests. If not it’s recommended that the user propagate ETR_TRACE_ID manually.

After the test execution etr will attempt to identify orphaned processes and perform cleanup as necessary. This is done by inspecting the environment of processes owned by the same user that invoked etr. This means that no cleanup is performed if process changes uid or modifies the environment to remove ETR_TRACE_ID.

When cleaning up etr will first send SIGTERM signal and if process does not exit it will send SIGKILL after 6 attemps with a 1 second delay between attempts.

Exit Codes¶

etr use the following exit codes:

EC	Description
0	Success.
126	Test runner reported failure.
127	Error occurred during execution. This includes transient errors such as test resources could not be fetched.
253	Aborted by user.
255	Internal error.