Merge ~afreiberger/charm-grafana:add_tests_and_linting into charm-grafana:master

[Drew Freiberger]

Added testing and resolved lint errors

[🤖 Canonical IS Merge Bot (canonical-is-mergebot)]

This merge proposal is being monitored by mergebot. Change the status to Approved to merge.

[🤖 Canonical IS Merge Bot (canonical-is-mergebot)]

Unable to determine commit message from repository - please click "Set commit message" and enter the commit message manually.

[🤖 Canonical IS Merge Bot (canonical-is-mergebot)]

This merge proposal is being monitored by mergebot. Change the status to Approved to merge.

[🤖 Canonical IS Merge Bot (canonical-is-mergebot)]

Unable to determine commit message from repository - please click "Set commit message" and enter the commit message manually.

[Peter Sabaini (peter-sabaini)]

Hi,

generally looking good. However I get two unit test errors, one is a left-over example, the other missing fixture:

file /home/peter/src/charms/grafana-charm/tests/unit/test_lib.py, line 8
      def test_grafana(self, grafana): E fixture 'grafana' not found > available fixtures: cache, capfd, capfdbinary, caplog, capsys, capsysbinary, cov, doctest_namespace, mock_charm_dir, mock_hookenv_config, mock_layers, mock_remote_unit, monkeypatch, no_cover, pytestconfig, record_property, rec
ord_testsuite_property, record_xml_attribute, recwarn, tmp_path, tmp_path_factory, tmpdir, tmpdir_factory
> use 'pytest --fixtures [testpath]' for help on them.

I've marked some more places with boilerplate code

cheers,
peter.

review: Needs Fixing

[Peter Sabaini (peter-sabaini)]

Forgot to note that the example-action test also failing in functional testing

review: Needs Fixing

[Drew Freiberger (afreiberger)]

Thanks for the review. I think I'll start with the unit_testing from guoqian's codebase. This was definitely WIP stuff I started friday afternoon and just wanted to get into the repo and test MRs with merge dependencies.

Thank you for the comments!

[Alvaro Uria (aluria)]

I have worked on adding tests starting from this branch. Since most of the code was a port from template-python-pytest, I have removed the unit tests (no real testing being done) and updated the functional tests following my work at [1].

I will go ahead and reject this change, and will propose a new fix soon.

1. https://bugs.launchpad.net/charm-grafana/+bug/1822329

review: Disapprove

Unmerged commits

d3370e1... by Drew Freiberger on 2020-01-11

Added testing and resolved lint errors

WIP

f2a62d8... by Drew Freiberger on 2020-01-10

Check dashboards before uploading new revisions

It was found that dashboards were being uploaded and creating unbounded
revision history once every 5 minutes during update-status causing the
grafana.db configuration database to balloon. To eliminate this, we
now validate that the rendered dashboard template result is not already
the version available in the grafana database before uploading.

Some refactoring of the dashboard function has been made to resolve
complexity warnings.

Closes-Bug: 1858490

diff --git a/.gitignore b/.gitignore
 new file mode 100644
index 0000000..32e2995
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1,22 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# Log files
+*.log
+
+.tox/
+.coverage
+
+# vi
+.*.swp
+
+# pycharm
+.idea/
+
+# version data
+repo-info
+
+# reports
+report/*
diff --git a/Makefile b/Makefile
 new file mode 100644
index 0000000..b357248
--- /dev/null
+++ b/Makefile
@@ -0,0 +1,49 @@
+help:
+	@echo "This project supports the following targets"
+	@echo ""
+	@echo " make help - show this text"
+	@echo " make submodules - make sure that the submodules are up-to-date"
+	@echo " make lint - run flake8"
+	@echo " make test - run the unittests and lint"
+	@echo " make unittest - run the tests defined in the unittest subdirectory"
+	@echo " make functional - run the tests defined in the functional subdirectory"
+	@echo " make release - build the charm"
+	@echo " make clean - remove unneeded files"
+	@echo ""
+
+submodules:
+	@echo "Cloning submodules"
+	@git submodule update --init --recursive
+
+lint:
+	@echo "Running flake8"
+	@tox -e lint
+
+test: lint unittest functional
+
+unittest:
+	@tox -e unit
+
+functional: build
+	@PYTEST_KEEP_MODEL=$(PYTEST_KEEP_MODEL) \
+	    PYTEST_CLOUD_NAME=$(PYTEST_CLOUD_NAME) \
+	    PYTEST_CLOUD_REGION=$(PYTEST_CLOUD_REGION) \
+	    tox -e functional
+
+build:
+	@echo "Building charm to base directory $(JUJU_REPOSITORY)"
+	@-git describe --tags > ./repo-info
+	@CHARM_LAYERS_DIR=./layers CHARM_INTERFACES_DIR=./interfaces TERM=linux \
+		JUJU_REPOSITORY=$(JUJU_REPOSITORY) charm build . --force
+
+release: clean build
+	@echo "Charm is built at $(JUJU_REPOSITORY)/builds"
+
+clean:
+	@echo "Cleaning files"
+	@if [ -d .tox ] ; then rm -r .tox ; fi
+	@if [ -d .pytest_cache ] ; then rm -r .pytest_cache ; fi
+	@find . -iname __pycache__ -exec rm -r {} +
+
+# The targets below don't depend on a file
+.PHONY: lint test unittest functional build release clean help submodules
diff --git a/lib/charms/layer/grafana.py b/lib/charms/layer/grafana.py
index b482203..67b53ea 100644
--- a/lib/charms/layer/grafana.py
+++ b/lib/charms/layer/grafana.py
@@ -2,13 +2,12 @@
 
 import json
 import requests
+from charmhelpers.core import unitdata
 from charmhelpers.core.hookenv import (
     config,
     log,
 )
 
-from charmhelpers.core import unitdata
-
 
 def get_admin_password():
     kv = unitdata.kv()
@@ -30,7 +29,7 @@ def import_dashboard(dashboard, name=None):
         name = dashboard['dashboard'].get('title') or 'Untitled'
     headers = {'Content-Type': 'application/json'}
     import_url = 'http://localhost:{}/api/dashboards/db'.format(
-                  config('port'))
+                 config('port'))
     passwd = get_admin_password()
     if passwd is None:
         return (False, 'Unable to retrieve grafana password.')
diff --git a/reactive/grafana.py b/reactive/grafana.py
index fbc53b8..3f24f0a 100644
--- a/reactive/grafana.py
+++ b/reactive/grafana.py
@@ -4,13 +4,11 @@ import glob
 import json
 import os
 import re
-import requests
 import shutil
-import six
 import subprocess
 import time
-from jsondiff import diff
 
+from charmhelpers import fetch
 from charmhelpers.contrib.charmsupport import nrpe
 from charmhelpers.core import (
     hookenv,
@@ -18,11 +16,9 @@ from charmhelpers.core import (
     unitdata,
 )
 from charmhelpers.core.templating import render
-from charmhelpers import fetch
-from charms.reactive.helpers import (
-    any_file_changed,
-    is_state,
-)
+
+from charms.layer import snap
+from charms.layer.grafana import import_dashboard
 from charms.reactive import (
     hook,
     remove_state,
@@ -30,11 +26,20 @@ from charms.reactive import (
     when,
     when_not,
 )
+from charms.reactive.helpers import (
+    any_file_changed,
+    is_state,
+)
 
-from charms.layer import snap
-from charms.layer.grafana import import_dashboard
 from jinja2 import Environment, FileSystemLoader, exceptions
 
+from jsondiff import diff
+
+import requests
+
+import six
+
+
 SVCNAME = {'snap': 'snap.grafana.grafana',
            'apt': 'grafana-server'}
 SNAP_NAME = 'grafana'
@@ -120,8 +125,7 @@ def install_packages():
         set_state('grafana.installed')
         hookenv.status_set('active', 'Completed installing grafana')
     elif source == 'snap' and \
-            (host.lsb_release()['DISTRIB_CODENAME'] >= 'xenial' or
-             host.lsb_release()['DISTRIB_CODENAME'] < 'p'):
+            (host.lsb_release()['DISTRIB_CODENAME'] >= 'xenial' or host.lsb_release()['DISTRIB_CODENAME'] < 'p'):
         # NOTE(aluria): precise is the last supported Ubuntu release, so
         # anything below 'p' is actually newer than xenial (systemd support)
         snap.install(SNAP_NAME, channel=channel, force_dangerous=False)
@@ -433,8 +437,9 @@ def configure_website(website):
 
 
 def validate_datasources():
-    """TODO: make sure datasources option is merged with
-    relation data
+    """Verify that datasources configuration is valid, if existing.
+
+    TODO: make sure datasources option is merged with relation data
     TODO: make sure datasources are validated
     """
     config = hookenv.config()
@@ -448,7 +453,8 @@ def validate_datasources():
 
 
 def check_datasource(ds):
-    """
+    """Check for and add datasources not currently in grafana DB.
+
     CREATE TABLE `data_source` (
     `id` INTEGER PRIMARY KEY AUTOINCREMENT NOT NULL
     , `org_id` INTEGER NOT NULL
@@ -470,7 +476,6 @@ def check_datasource(ds):
     , `with_credentials` INTEGER NOT NULL DEFAULT 0);
     INSERT INTO "data_source" VALUES(1,1,0,'prometheus','BootStack Prometheus','proxy','http://localhost:9090','','','',0,'','',1,'{}','2016-01-22 12:11:06','2016-01-22 12:11:11',0);
     """  # noqa E501
-
     # ds will be similar to:
     # {'service_name': 'prometheus',
     #  'url': 'http://10.0.3.216:9090',
@@ -505,9 +510,8 @@ def check_datasource(ds):
 
 # This isn't exposed in charmhelpers: https://github.com/juju/charm-helpers/issues/367
 def render_custom(source, context, **parameters):
-    """
-    Renders a template from the template folder with custom environment
-    parameters.
+    """Render a template from the template folder with custom environment parameters.
+
     source: template file name to render from
     context: template context variables
     parameters: initialization parameters for the jinja Environment
@@ -703,7 +707,8 @@ def generate_query(ds, is_default, id=None):
 @when('grafana.started')
 @when_not('grafana.admin_password.set')
 def check_adminuser():
-    """
+    """Create Adminuser if not existing.
+
     CREATE TABLE `user` (
     `id` INTEGER PRIMARY KEY AUTOINCREMENT NOT NULL
     , `version` INTEGER NOT NULL
@@ -723,7 +728,6 @@ def check_adminuser():
     );
     INSERT INTO "user" VALUES(1,0,'admin','[e-mail]','BootStack Team','309bc4e78bc60d02dc0371d9e9fa6bf9a809d5dc25c745b9e3f85c3ed49c6feccd4ffc96d1db922f4297663a209e93f7f2b6','LZeJ3nSdrC','hseJcLcnPN','',1,1,0,'light','2016-01-22 12:00:08','2016-01-22 12:02:13');
     """  # noqa E501
-
     # XXX: If you add any dependencies on config items here,
     #      be sure to update config_changed() accordingly!
 
@@ -759,9 +763,7 @@ def check_adminuser():
         query = cur.execute('SELECT id, login, salt FROM user')
         for row in query.fetchall():
             if row[1] == 'admin':
-                nagios_context = config.get('nagios_context', False)
-                if not nagios_context:
-                    nagios_context = 'UNKNOWN'
+                nagios_context = config.get('nagios_context', 'UNKNOWN')
                 email = '[e-mail]' % nagios_context
                 hpasswd = hpwgen(passwd, row[2])
                 if hpasswd:
diff --git a/requirements.txt b/requirements.txt
 new file mode 100644
index 0000000..8462291
--- /dev/null
+++ b/requirements.txt
@@ -0,0 +1 @@
+# Include python requirements here
diff --git a/tests/functional/__pycache__/conftest.cpython-37-pytest-5.3.2.pyc b/tests/functional/__pycache__/conftest.cpython-37-pytest-5.3.2.pyc
 new file mode 100644
index 0000000..54c4d14
 Binary files /dev/null and b/tests/functional/__pycache__/conftest.cpython-37-pytest-5.3.2.pyc differ
diff --git a/tests/functional/__pycache__/juju_tools.cpython-37.pyc b/tests/functional/__pycache__/juju_tools.cpython-37.pyc
 new file mode 100644
index 0000000..ce8ffdc
 Binary files /dev/null and b/tests/functional/__pycache__/juju_tools.cpython-37.pyc differ
diff --git a/tests/functional/__pycache__/test_deploy.cpython-37-pytest-5.3.2.pyc b/tests/functional/__pycache__/test_deploy.cpython-37-pytest-5.3.2.pyc
 new file mode 100644
index 0000000..27b64d7
 Binary files /dev/null and b/tests/functional/__pycache__/test_deploy.cpython-37-pytest-5.3.2.pyc differ
diff --git a/tests/functional/conftest.py b/tests/functional/conftest.py
 new file mode 100644
index 0000000..1d2d6e9
--- /dev/null
+++ b/tests/functional/conftest.py
@@ -0,0 +1,67 @@
+#!/usr/bin/python3
+"""
+Reusable pytest fixtures for functional testing
+
+Environment variables
+---------------------
+
+PYTEST_CLOUD_REGION, PYTEST_CLOUD_NAME: cloud name and region to use for juju model creation
+
+PYTEST_KEEP_MODEL: if set, the testing model won't be torn down at the end of the testing session
+
+"""
+
+import asyncio
+import os
+import uuid
+import pytest
+import subprocess
+
+from juju.controller import Controller
+from juju_tools import JujuTools
+
+
[e-mail](scope='module')
+def event_loop():
+    """Override the default pytest event loop to allow for fixtures using a broader scope"""
+    loop = asyncio.get_event_loop_policy().new_event_loop()
+    asyncio.set_event_loop(loop)
+    loop.set_debug(True)
+    yield loop
+    loop.close()
+    asyncio.set_event_loop(None)
+
+
[e-mail](scope='module')
+async def controller():
+    """Connect to the current controller"""
+    _controller = Controller()
+    await _controller.connect_current()
+    yield _controller
+    await _controller.disconnect()
+
+
[e-mail](scope='module')
+async def model(controller):
+    """This model lives only for the duration of the test"""
+    model_name = "functest-{}".format(str(uuid.uuid4())[-12:])
+    _model = await controller.add_model(model_name,
+                                        cloud_name=os.getenv('PYTEST_CLOUD_NAME'),
+                                        region=os.getenv('PYTEST_CLOUD_REGION'),
+                                        )
+    # https://github.com/juju/python-libjuju/issues/267
+    subprocess.check_call(['juju', 'models'])
+    while model_name not in await controller.list_models():
+        await asyncio.sleep(1)
+    yield _model
+    await _model.disconnect()
+    if not os.getenv('PYTEST_KEEP_MODEL'):
+        await controller.destroy_model(model_name)
+        while model_name in await controller.list_models():
+            await asyncio.sleep(1)
+
+
[e-mail](scope='module')
+async def jujutools(controller, model):
+    tools = JujuTools(controller, model)
+    return tools
diff --git a/tests/functional/juju_tools.py b/tests/functional/juju_tools.py
 new file mode 100644
index 0000000..4b4884f
--- /dev/null
+++ b/tests/functional/juju_tools.py
@@ -0,0 +1,68 @@
+import pickle
+import juju
+import base64
+
+# from juju.errors import JujuError
+
+
+class JujuTools:
+    def __init__(self, controller, model):
+        self.controller = controller
+        self.model = model
+
+    async def run_command(self, cmd, target):
+        """
+        Runs a command on a unit.
+
+        :param cmd: Command to be run
+        :param unit: Unit object or unit name string
+        """
+        unit = (
+            target
+            if isinstance(target, juju.unit.Unit)
+            else await self.get_unit(target)
+        )
+        action = await unit.run(cmd)
+        return action.results
+
+    async def remote_object(self, imports, remote_cmd, target):
+        """
+        Runs command on target machine and returns a python object of the result
+
+        :param imports: Imports needed for the command to run
+        :param remote_cmd: The python command to execute
+        :param target: Unit object or unit name string
+        """
+        python3 = "python3 -c '{}'"
+        python_cmd = ('import pickle;'
+                      'import base64;'
+                      '{}'
+                      'print(base64.b64encode(pickle.dumps({})), end="")'
+                      .format(imports, remote_cmd))
+        cmd = python3.format(python_cmd)
+        results = await self.run_command(cmd, target)
+        return pickle.loads(base64.b64decode(bytes(results['Stdout'][2:-1], 'utf8')))
+
+    async def file_stat(self, path, target):
+        """
+        Runs stat on a file
+
+        :param path: File path
+        :param target: Unit object or unit name string
+        """
+        imports = 'import os;'
+        python_cmd = ('os.stat("{}")'
+                      .format(path))
+        print("Calling remote cmd: " + python_cmd)
+        return await self.remote_object(imports, python_cmd, target)
+
+    async def file_contents(self, path, target):
+        """
+        Returns the contents of a file
+
+        :param path: File path
+        :param target: Unit object or unit name string
+        """
+        cmd = 'cat {}'.format(path)
+        result = await self.run_command(cmd, target)
+        return result['Stdout']
diff --git a/tests/functional/requirements.txt b/tests/functional/requirements.txt
 new file mode 100644
index 0000000..f76bfbb
--- /dev/null
+++ b/tests/functional/requirements.txt
@@ -0,0 +1,6 @@
+flake8
+juju
+mock
+pytest
+pytest-asyncio
+requests
diff --git a/tests/functional/test_deploy.py b/tests/functional/test_deploy.py
 new file mode 100644
index 0000000..bf28c17
--- /dev/null
+++ b/tests/functional/test_deploy.py
@@ -0,0 +1,103 @@
+import os
+import pytest
+import subprocess
+import stat
+
+# Treat all tests as coroutines
+pytestmark = pytest.mark.asyncio
+
+juju_repository = os.getenv('JUJU_REPOSITORY', '.').rstrip('/')
+series = ['xenial',
+          'bionic',
+          pytest.param('eoan', marks=pytest.mark.xfail(reason='canary')),
+          ]
+sources = [('local', '{}/builds/grafana'.format(juju_repository)),
+           # ('jujucharms', 'cs:...'),
+           ]
+
+
+# Uncomment for re-using the current model, useful for debugging functional tests
+# @pytest.fixture(scope='module')
+# async def model():
+#     from juju.model import Model
+#     model = Model()
+#     await model.connect_current()
+#     yield model
+#     await model.disconnect()
+
+
+# Custom fixtures
[e-mail](params=series)
+def series(request):
+    return request.param
+
+
[e-mail](params=sources, ids=[s[0] for s in sources])
+def source(request):
+    return request.param
+
+
[e-mail]
+async def app(model, series, source):
+    app_name = 'grafana-{}-{}'.format(series, source[0])
+    return await model._wait_for_new('application', app_name)
+
+
+async def test_grafana_deploy(model, series, source, request):
+    # Starts a deploy for each series
+    # Using subprocess b/c libjuju fails with JAAS
+    # https://github.com/juju/python-libjuju/issues/221
+    application_name = 'grafana-{}-{}'.format(series, source[0])
+    cmd = ['juju', 'deploy', source[1], '-m', model.info.name,
+           '--series', series, application_name]
+    if request.node.get_closest_marker('xfail'):
+        # If series is 'xfail' force install to allow testing against versions not in
+        # metadata.yaml
+        cmd.append('--force')
+    subprocess.check_call(cmd)
+
+
+async def test_charm_upgrade(model, app):
+    if app.name.endswith('local'):
+        pytest.skip("No need to upgrade the local deploy")
+    unit = app.units[0]
+    await model.block_until(lambda: unit.agent_status == 'idle')
+    subprocess.check_call(['juju',
+                           'upgrade-charm',
+                           '--switch={}'.format(sources[0][1]),
+                           '-m', model.info.name,
+                           app.name,
+                           ])
+    await model.block_until(lambda: unit.agent_status == 'executing')
+
+
+# Tests
+async def test_grafana_status(model, app):
+    # Verifies status for all deployed series of the charm
+    await model.block_until(lambda: app.status == 'active')
+    unit = app.units[0]
+    await model.block_until(lambda: unit.agent_status == 'idle')
+
+
+async def test_example_action(app):
+    unit = app.units[0]
+    action = await unit.run_action('example-action')
+    action = await action.wait()
+    assert action.status == 'completed'
+
+
+async def test_run_command(app, jujutools):
+    unit = app.units[0]
+    cmd = 'hostname -i'
+    results = await jujutools.run_command(cmd, unit)
+    assert results['Code'] == '0'
+    assert unit.public_address in results['Stdout']
+
+
+async def test_file_stat(app, jujutools):
+    unit = app.units[0]
+    path = '/var/lib/juju/agents/unit-{}/charm/metadata.yaml'.format(unit.entity_id.replace('/', '-'))
+    fstat = await jujutools.file_stat(path, unit)
+    assert stat.filemode(fstat.st_mode) == '-rw-r--r--'
+    assert fstat.st_uid == 0
+    assert fstat.st_gid == 0
diff --git a/tests/unit/__pycache__/conftest.cpython-37-pytest-5.3.2.pyc b/tests/unit/__pycache__/conftest.cpython-37-pytest-5.3.2.pyc
 new file mode 100644
index 0000000..f22ae79
 Binary files /dev/null and b/tests/unit/__pycache__/conftest.cpython-37-pytest-5.3.2.pyc differ
diff --git a/tests/unit/__pycache__/test_actions.cpython-37-pytest-5.3.2.pyc b/tests/unit/__pycache__/test_actions.cpython-37-pytest-5.3.2.pyc
 new file mode 100644
index 0000000..dec8696
 Binary files /dev/null and b/tests/unit/__pycache__/test_actions.cpython-37-pytest-5.3.2.pyc differ
diff --git a/tests/unit/__pycache__/test_lib.cpython-37-pytest-5.3.2.pyc b/tests/unit/__pycache__/test_lib.cpython-37-pytest-5.3.2.pyc
 new file mode 100644
index 0000000..0ef154b
 Binary files /dev/null and b/tests/unit/__pycache__/test_lib.cpython-37-pytest-5.3.2.pyc differ
diff --git a/tests/unit/conftest.py b/tests/unit/conftest.py
 new file mode 100644
index 0000000..a2a5470
--- /dev/null
+++ b/tests/unit/conftest.py
@@ -0,0 +1,69 @@
+#!/usr/bin/python3
+import mock
+import pytest
+
+
+# If layer options are used, add this to ${fixture}
+# and import layer in ${libfile}
[e-mail]
+def mock_layers(monkeypatch):
+    import sys
+    sys.modules['charms.layer'] = mock.Mock()
+    sys.modules['reactive'] = mock.Mock()
+    # Mock any functions in layers that need to be mocked here
+
+    def options(layer):
+        # mock options for layers here
+        if layer == 'example-layer':
+            options = {'port': 9999}
+            return options
+        else:
+            return None
+
+    monkeypatch.setattr('${libfile}.layer.options', options)
+
+
[e-mail]
+def mock_hookenv_config(monkeypatch):
+    import yaml
+
+    def mock_config():
+        cfg = {}
+        yml = yaml.load(open('./config.yaml'))
+
+        # Load all defaults
+        for key, value in yml['options'].items():
+            cfg[key] = value['default']
+
+        # Manually add cfg from other layers
+        # cfg['my-other-layer'] = 'mock'
+        return cfg
+
+    monkeypatch.setattr('${libfile}.hookenv.config', mock_config)
+
+
[e-mail]
+def mock_remote_unit(monkeypatch):
+    monkeypatch.setattr('${libfile}.hookenv.remote_unit', lambda: 'unit-mock/0')
+
+
[e-mail]
+def mock_charm_dir(monkeypatch):
+    monkeypatch.setattr('${libfile}.hookenv.charm_dir', lambda: '/mock/charm/dir')
+
+
+# @pytest.fixture
+# def ${fixture}(tmpdir, mock_hookenv_config, mock_charm_dir, monkeypatch):
+#     from $libfile import $libclass
+#     helper = ${libclass}()
+#
+#     # Example config file patching
+#     cfg_file = tmpdir.join('example.cfg')
+#     with open('./tests/unit/example.cfg', 'r') as src_file:
+#         cfg_file.write(src_file.read())
+#     helper.example_config_file = cfg_file.strpath
+#
+#     # Any other functions that load helper will get this version
+#     monkeypatch.setattr('${libfile}.${libclass}', lambda: helper)
+#
+#     return helper
diff --git a/tests/unit/example.cfg b/tests/unit/example.cfg
 new file mode 100644
index 0000000..81b1e94
--- /dev/null
+++ b/tests/unit/example.cfg
@@ -0,0 +1 @@
+This is an example config file included with the unit tests
diff --git a/tests/unit/requirements.txt b/tests/unit/requirements.txt
 new file mode 100644
index 0000000..9c685e5
--- /dev/null
+++ b/tests/unit/requirements.txt
@@ -0,0 +1,5 @@
+charmhelpers
+charms.reactive
+mock
+pytest
+pytest-cov
diff --git a/tests/unit/test_actions.py b/tests/unit/test_actions.py
 new file mode 100644
index 0000000..6e7cddb
--- /dev/null
+++ b/tests/unit/test_actions.py
@@ -0,0 +1,12 @@
+import imp
+
+import mock
+
+
+class TestActions():
+    def test_example_action(self, my_action, monkeypatch):
+        mock_function = mock.Mock()
+        monkeypatch.setattr(my_action, 'action_function', mock_function)
+        assert mock_function.call_count == 0
+        imp.load_source('action_function', './actions/example-action')
+        assert mock_function.call_count == 1
diff --git a/tests/unit/test_lib.py b/tests/unit/test_lib.py
 new file mode 100644
index 0000000..a7b2b08
--- /dev/null
+++ b/tests/unit/test_lib.py
@@ -0,0 +1,12 @@
+#!/usr/bin/python3
+
+
+class TestLib():
+    def test_pytest(self):
+        assert True
+
+    def test_grafana(self, grafana):
+        """See if the helper fixture works to load charm configs."""
+        assert isinstance(grafana.charm_config, dict)
+
+    # Include tests for functions in ${libfile}
diff --git a/tox.ini b/tox.ini
 new file mode 100644
index 0000000..8b4adc3
--- /dev/null
+++ b/tox.ini
@@ -0,0 +1,71 @@
+[tox]
+skipsdist=True
+envlist = unit, functional
+skip_missing_interpreters = True
+
+[testenv]
+basepython = python3
+setenv =
+  PYTHONPATH = .
+
+[testenv:unit]
+commands = pytest -v --ignore {toxinidir}/tests/functional \
+	   --cov=lib \
+	   --cov=reactive \
+	   --cov=actions \
+	   --cov-report=term \
+	   --cov-report=annotate:report/annotated \
+	   --cov-report=html:report/html
+deps = -r{toxinidir}/tests/unit/requirements.txt
+       -r{toxinidir}/requirements.txt
+setenv = PYTHONPATH={toxinidir}/lib
+
+[testenv:functional]
+passenv =
+  HOME
+  JUJU_REPOSITORY
+  PATH
+  PYTEST_KEEP_MODEL
+  PYTEST_CLOUD_NAME
+  PYTEST_CLOUD_REGION
+commands = pytest -v --ignore {toxinidir}/tests/unit
+deps = -r{toxinidir}/tests/functional/requirements.txt
+       -r{toxinidir}/requirements.txt
+
+[testenv:lint]
+commands = flake8
+deps =
+    flake8
+    flake8-docstrings
+    flake8-import-order
+    pep8-naming
+    flake8-colors
+
+[flake8]
+exclude =
+    .git,
+    __pycache__,
+    .tox,
+# H405: Multi line docstrings should start with a one line summary followed by
+#       an empty line.
+# D100: Missing docstring in public module
+# D101: Missing docstring in public class
+# D102: Missing docstring in public method
+# D103: Missing docstring in public function
+# D104: Missing docstring in public package
+# D105: Missing docstring in magic method
+# D107: Missing docstring in __init__
+# D200: One-line docstring should fit on one line with quotes
+# D202: No blank lines allowed after function docstring
+# D203: 1 blank required before class docstring
+# D204: 1 blank line required after class docstring
+# D205: 1 blank line required between summary line and description
+# D208: Docstring is over-indented
+# D400: First line should end with a period
+# D401: First line should be in imperative mood
+# I201: Missing newline between import groups
+# I100: Import statements are in the wrong order
+
+ignore = H405,D100,D101,D102,D103,D104,D105,D107,D200,D202,D203,D204,D205,D208,D400,D401,I100,I201
+max-line-length = 120
+max-complexity = 10