How to reuse ssh connection during entire pytest suite - ssh

Question

I have an e2e test suite which loads some fixtures in in the database by calling a script on the server side using an SSH connection.
I want to keep the fixtures that i load local to test that needs them. I would write a test something like
class ExampleTests(BaseTest):
def test_A(self):
load_fixture('TEST_A')
do_actual_test()
def test_B(self):
load_fixture('TEST_B')
do_actual_test()
In my load_fixture method the SSH connection is made and the script is run on the server side.
If i run the entire test suite it will create a new SSH connection each time I call the load_fixture method. Conceptually this is what i want. I don't want to load all my fixtures for all my tests before any test runs. I want to be able to run fixtures when i need them. e.g.
class ExampleTests(BaseTest):
def test_B(self):
user_a = load_user_fixture('username-A')
do_some_testing_on_user_a()
load_post_fixture_for_user(user_a, subject='subject-a')
do_tests_using_post()
In this test it would also create 2 ssh connections.
So what i want to have happen is that the first time i call the load_fixture method it creates the connection but keeps it around for the duration of the test suite. Or i create a connection before any test runs and then use that connection whenever i load a fixture.
Of course it should keep working when i run the tests over multiple core.
My load_fixture function looks something like:
def load_fixtures(connection_info, command, fixtures):
out, err, exit_code = run_remote_fixture_script(connection_info, command, fixtures)
def run_remote_fixture_script(connection_info, command_name, *args):
ssh = SSHClient()
ssh.connect(...)
command = '''
./load_fixture_script {test_target} {command} {args};
'''.format(
test_target=connection_info.target,
command=command_name,
args=''.join([" '{}'".format(arg) for arg in args])
)
stdin, stdout, stderr = ssh.exec_command(command)
exit_code = stdout.channel.recv_exit_status()
ssh.close()
return stdout, stderr, exit_code
I also want to reopen the connection automatically if for any reason the connection closes.

Answer 1

You need to use
#pytest.fixture(scope="module")
Keeping the scope as module will keep it for whole test suite.
and finalizer method within your fixture
`def run_remote_fixture_script(connection_info, command_name, *args):
ssh = SSHClient()
ssh.connect(...)
command = '''
./load_fixture_script {test_target} {command} {args};
'''.format(
test_target=connection_info.target,
command=command_name,
args=''.join([" '{}'".format(arg) for arg in args])
)
stdin, stdout, stderr = ssh.exec_command(command)
exit_code = stdout.channel.recv_exit_status()
def fin():
print ("teardown ssh")
ssh.close()
request.addfinalizer(fin)
return stdout, stderr, exit_code
Please Excuse the formating of code. You could see this link for more details
And you would call this fixture as
def test_function(run_remote_fixture_script)
output = run_remote_fixture_script
Hope this helps .
Finalizer method will be called end of test suite , if scope is method it will be called after method