91.0 Thelma & Louise

The Avocado team is proud to present another release: Avocado 91.0, AKA “Thelma & Louise”, is now available!

Release documentation: Avocado 91.0

Important Announcement

Avocado has switched the default runner, from the implementation most people currently use (internally simply called runner), to the newer architecture and implementation called nrunner.

Users installing and relying on the latest Avocado release will be impacted by this change and should act accordingly.

To keep using the current (soon to be legacy) runner, you must set the --test-runner=runner command line option (or the equivalent test_runner configuration option, under section [run]).

Known issues are being tracked on our GitHub project page, with the nrunner tag, and new issue reports are appreciated.

Users/Test Writers

  • As per the previous section, the nrunner test runner implementation is now the default on every avocado run command (or equivalent Job API scripts). Since the previous release, nrunner supports:

  1. the “fail fast” (run --failfast) feature.

  2. the varianter feature.

  3. UNIX domain sockets as the communication channel with runners (the new default).

  4. a sysinfo runner, which will allow for sysinfo collection on any supported spawner.

  5. early notification of missing runners for tasks in the requested suite.

  • The assets plugin fetch command (avocado assets fetch) now supports:

  1. fetching assets defined in a Python list in INSTRUMENTED tests.

  2. setting a timeout for the download of assets.

  • Improved checks when users attempt to use the varianter and simple parameters (-p) at the same time.

  • The Podman spawner (--nrunner-spawner=podman) will now attempt to use a Container image (--spawner-podman-image=) that matches the host Linux distribution. If it’s not possible to detect the host distribution, the latest Fedora image will be used.

Bug Fixes

  • The extraction of DEB packages by means of avocado.utils.software_manager.SoftwareManager.extract_from_package() was fixed and does not depend on ar utility anymore (as it now uses the avocado.utils.ar module).

  • The --store-logging-stream parameter value was being incorrectly parsed as a list of characters. If a bar value is given, it would generate the b.INFO, a.INFO, and r.INFO file. The fix parses the command line arguments by treating the value as a comma separated list (that becomes a set).

  • nrunner will now properly translate reference names with absolute paths into Python unittest “dotted names”.

  • The TAP parser (avocado.core.tapparser) will not choke on unexpected content, ignoring it according to the standard.

  • avocado.core.nrunner.Runnables created by suites will now contain the full suite configuration.

  • If a job contains multiple test suites with the same name, and tests within those suites also have the same name, test results would be overwritten. Now job name uniqueness is enforced and no test results from a suite should be able to overwrite other’s.

  • Some occurrences of the incorrect AVOCADO_TEST_OUTPUT_DIR environment variable name were renamed to the proper name (AVOCADO_TEST_OUTPUTDIR).

Utility APIs

  • avocado.utils.network.interfaces.NetworkInterface can now access and present information on interfaces that do not have an IP address assigned to them.

  • avocado.utils.distro can now detect the distribution on remote machines.

  • A new avocado.utils.ar module was introduced that allows extraction of UNIX ar archive contents.

  • A new avocado.utils.sysinfo module that powers the sysinfo feature, but is now also accessible to job/test writers.

  • Times related to the duration of tasks are now limited to nanosecond precision to improve readability.

  • The avocado.utils.distro will now correctly return a avocado.utils.distro.UNKNOWN_DISTRO on non UNIX systems, instead of crashing.

  • The avocado.utils.network.hosts won’t consider anymore bonding_masters, a file that may exist at /sys/class/net, as the name of an interface.

Misc Changes

  • Many documentation improvements.

  • The setup.py script received many improvements, including:

  1. A new test command implementation.

  2. A new plugin command.

  • Various logging related improvements, especially regarding the elimination of Python’s root logger usage.

Internal Changes

  • Major improvements to the CI, including:

  1. additional tests for setup.py

  2. most of the “pre-release” test plan was migrated to a CI job

  • The avocado.Test.workdir() is now initialized lazily, which prevents never used work directories from being created and kept in the test results’ directory.

  • A circular import condition was fixed related to the Settings plugins initialized during avocado early initialization.

  • A rename of the requirements-selftests.txt file to requirements-dev.txt to better capture the fact that the selftests requirements are a subset of the requirements needed for development.

  • selftests/safeloader.sh now supports checking a single file.

  • Debugging aids were added to the Task/StateMachine/Repo systems of the nrunner. These can be seen by default at the avocado.core.DEBUG file in the job’s result directory.

  • An nrunner based job its tasks are now better bound by the job ID, with messages being logged in case the status server receives messages destined for another job.

  • The very long and detailed temporary directory prefix used on TestCaseTmpDir was shortened so that limitations on paths such as socket names are not easily hit.

  • Clean up of duplicate methods in the avocado.core.loader.FileLoader class.

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