xdist: pytest distributed testing plugin
The pytest-xdist plugin extends pytest with some unique test execution modes:
- test run parallelization: if you have multiple CPUs or hosts you can use those for a combined test run. This allows to speed up development or to use special resources of remote machines.
--looponfail: run your tests repeatedly in a subprocess. After each run pytest waits until a file in your project changes and then re-runs the previously failing tests. This is repeated until all tests pass after which again a full run is performed.
- Multi-Platform coverage: you can specify different Python interpreters or different platforms and run tests in parallel on all of them.
Before running tests remotely,
pytest efficiently "rsyncs" your
program source code to the remote place. All test results
are reported back and displayed to your local terminal.
You may specify different Python versions and interpreters.
If you would like to know how pytest-xdist works under the covers, checkout OVERVIEW.
Install the plugin with:
pip install pytest-xdist
or use the package in develop/in-place mode with a checkout of the pytest-xdist repository
pip install --editable .
Speed up test runs by sending tests to multiple CPUs
To send tests to multiple CPUs, type:
pytest -n NUM
Especially for longer running tests or tests requiring
a lot of I/O this can lead to considerable speed ups. This option can
also be set to
auto for automatic detection of the number of CPUs.
If a test crashes the interpreter, pytest-xdist will automatically restart
that worker and report the failure as usual. You can use the
--max-worker-restart option to limit the number of workers that can
be restarted, or disable restarting altogether using
By default, the
-n option will send pending tests to any worker that is available, without
any guaranteed order, but you can control this with these options:
--dist=loadscope: tests will be grouped by module for test functions and by class for test methods, then each group will be sent to an available worker, guaranteeing that all tests in a group run in the same process. This can be useful if you have expensive module-level or class-level fixtures. Currently the groupings can't be customized, with grouping by class takes priority over grouping by module. This feature was added in version
--dist=loadfile: tests will be grouped by file name, and then will be sent to an available worker, guaranteeing that all tests in a group run in the same worker. This feature was added in version
Making session-scoped fixtures execute only once
pytest-xdist is designed so that each worker process will perform its own collection and execute
a subset of all tests. This means that tests in different processes requesting a high-level
scoped fixture (for example
session) will execute the fixture code more than once, which
breaks expectations and might be undesired in certain situations.
pytest-xdist does not have a builtin support for ensuring a session-scoped fixture is
executed exactly once, this can be achieved by using a lock file for inter-process communication.
The example below needs to execute the fixture
session_data only once (because it is
resource intensive, or needs to execute only once to define configuration options, etc), so it makes
use of a FileLock to produce the fixture data only once
when the first process requests the fixture, while the other processes will then read
the data from a file.
Here is the code:
import json import pytest from filelock import FileLock @pytest.fixture(scope="session") def session_data(tmp_path_factory, worker_id): if not worker_id: # not executing in with multiple workers, just produce the data and let # pytest's fixture caching do its job return produce_expensive_data() # get the temp directory shared by all workers root_tmp_dir = tmp_path_factory.getbasetemp().parent fn = root_tmp_dir / "data.json" with FileLock(str(fn) + ".lock"): if fn.is_file(): data = json.loads(fn.read_text()) else: data = produce_expensive_data() fn.write_text(json.dumps(data)) return data
The example above can also be use in cases a fixture needs to execute exactly once per test session, like initializing a database service and populating initial tables.
This technique might not work for every case, but should be a starting point for many situations where executing a high-scope fixture exactly once is important.
Running tests in a Python subprocess
To instantiate a python3.5 subprocess and send tests to it, you may type:
pytest -d --tx popen//python=python3.5
This will start a subprocess which is run with the
Python interpreter, found in your system binary lookup path.
If you prefix the --tx option value like this:
then three subprocesses would be created and tests will be load-balanced across these three processes.
Running tests in a boxed subprocess
This functionality has been moved to the
pytest-forked plugin, but the
is still kept for backward compatibility.
Sending tests to remote SSH accounts
Suppose you have a package
mypkg which contains some
tests that you can successfully run locally. And you
have a ssh-reachable machine
you can ad-hoc distribute your tests by typing:
pytest -d --tx ssh=myhostpopen --rsyncdir mypkg mypkg
This will synchronize your
mypkg package directory
to a remote ssh account and then locally collect tests
and send them to remote places for execution.
You can specify multiple
to be sent to the remote side.
For pytest to collect and send tests correctly
you not only need to make sure all code and tests
directories are rsynced, but that any test (sub) directory
also has an
__init__.py file because internally
pytest references tests as a fully qualified python
module path. You will otherwise get strange errors
during setup of the remote side.
You can specify multiple
--rsyncignore glob patterns
to be ignored when file are sent to the remote side.
There are also internal ignores:
.*, *.pyc, *.pyo, *~
Those you cannot override using rsyncignore command-line or
Sending tests to remote Socket Servers
Download the single-module socketserver.py Python program and run it like this:
It will tell you that it starts listening on the default port. You can now on your home machine specify this new socket host with something like this:
pytest -d --tx socket=192.168.1.102:8888 --rsyncdir mypkg mypkg
Running tests on many platforms at once
The basic command to run tests on multiple platforms is:
pytest --dist=each --tx=spec1 --tx=spec2
If you specify a windows host, an OSX host and a Linux environment this command will send each tests to all platforms - and report back failures from all platforms at once. The specifications strings use the xspec syntax.
Identifying the worker process during a test
New in version 1.15.
If you need to determine the identity of a worker process in
a test or fixture, you may use the
worker_id fixture to do so:
@pytest.fixture() def user_account(worker_id): """ use a different account in each xdist worker """ return "account_%s" % worker_id
xdist is disabled (running with
-n0 for example), then
worker_id will return
Additionally, worker processes have the following environment variables defined:
PYTEST_XDIST_WORKER: the name of the worker, e.g.,
PYTEST_XDIST_WORKER_COUNT: the total number of workers in this session, e.g.,
-n 4is given in the command-line.
The information about the worker_id in a test is stored in the
well, under the
sys.argv from the master node in workers
To access the
sys.argv passed to the command-line of the master node, use
Specifying test exec environments in an ini file
You can use pytest's ini file configuration to avoid typing common options. You can for example make running with three subprocesses your default like this:
[pytest] addopts = -n3
You can also add default environments like this:
[pytest] addopts = --tx ssh=myhost//python=python3.5 --tx ssh=myhost//python=python3.6
and then just type:
to run tests in each of the environments.
Specifying "rsync" dirs in an ini-file
setup.cfg file in your root project directory
you may specify directories to include or to exclude in synchronisation:
[pytest] rsyncdirs = . mypkg helperpkg rsyncignore = .hg
These directory specifications are relative to the directory where the configuration file was found.