106.0 - Taking Off

The Avocado team is proud to present another release: Avocado 106.0, AKA “Taking Off”, is now available!

Release documentation: Avocado 106.0

Users/Test Writers

The usage of Coverage.py (the de-facto Python tool for for test coverage) with Avocado received a number of improvements such as support for python-unittest tests in addition to avocado-instrumented tests.
When listing tests with avocado -V list, the resolver that handled the reference to test resolution is now shown. The reason is that with the introduction of resolvers such as runnable-recipe and runnables-recipe the resolver may not be the same as the test type.
Job log files can now use a memory buffer to reduce I/O pressure when there’s much log activity. Users can set the job.run.logging_buffer_size with the desired buffer size.

avocado.utils.disk.get_io_scheduler_list(): new utility function that returns the I/O scheduler available for the I/O device.
avocado.utils.disk.get_io_scheduler(): new utility function that returns the I/O scheduler which is currently set for a device.

The resolver could fail to check file permissions on systems such as macOS due to limitations in os.access(). A more robust check is now used based on os.stat().
A hardcoded reference and check for the “nrunner” runner plugin was removed.
A regression that was adding the entire Avocado configuration to every runnable was fixed.
The suite configuration is now applied to all runnables at suite creation time, instead of at just before the test execution time with the avocado run command/plugin.
The current configuration of a runnable is used to set its identifier.

For more information, please check out the complete Avocado changelog.

For more information on the actual issues addressed, please check out the milestone information.

For more information on the release codename, please refer to IMDb.