diff options
Diffstat (limited to 'roles/openshift_health_checker/test')
26 files changed, 3990 insertions, 0 deletions
diff --git a/roles/openshift_health_checker/test/action_plugin_test.py b/roles/openshift_health_checker/test/action_plugin_test.py new file mode 100644 index 000000000..c109ebd24 --- /dev/null +++ b/roles/openshift_health_checker/test/action_plugin_test.py @@ -0,0 +1,251 @@ +import pytest + +from ansible.playbook.play_context import PlayContext + +from openshift_health_check import ActionModule, resolve_checks +from openshift_checks import OpenShiftCheckException + + +def fake_check(name='fake_check', tags=None, is_active=True, run_return=None, run_exception=None, changed=False): + """Returns a new class that is compatible with OpenShiftCheck for testing.""" + + _name, _tags = name, tags + + class FakeCheck(object): + name = _name + tags = _tags or [] + changed = False + + def __init__(self, execute_module=None, task_vars=None, tmp=None): + pass + + def is_active(self): + return is_active + + def run(self): + self.changed = changed + if run_exception is not None: + raise run_exception + return run_return + + return FakeCheck + + +# Fixtures + + +@pytest.fixture +def plugin(): + task = FakeTask('openshift_health_check', {'checks': ['fake_check']}) + plugin = ActionModule(task, None, PlayContext(), None, None, None) + return plugin + + +class FakeTask(object): + def __init__(self, action, args): + self.action = action + self.args = args + self.async = 0 + + +@pytest.fixture +def task_vars(): + return dict(openshift=dict(), ansible_host='unit-test-host') + + +# Assertion helpers + + +def failed(result, msg_has=None): + if msg_has is not None: + assert 'msg' in result + for term in msg_has: + assert term.lower() in result['msg'].lower() + return result.get('failed', False) + + +def changed(result): + return result.get('changed', False) + + +# tests whether task is skipped, not individual checks +def skipped(result): + return result.get('skipped', False) + + +# Tests + + +@pytest.mark.parametrize('task_vars', [ + None, + {}, +]) +def test_action_plugin_missing_openshift_facts(plugin, task_vars, monkeypatch): + monkeypatch.setattr('openshift_health_check.resolve_checks', lambda *args: ['fake_check']) + result = plugin.run(tmp=None, task_vars=task_vars) + + assert failed(result, msg_has=['openshift_facts']) + + +def test_action_plugin_cannot_load_checks_with_the_same_name(plugin, task_vars, monkeypatch): + FakeCheck1 = fake_check('duplicate_name') + FakeCheck2 = fake_check('duplicate_name') + checks = [FakeCheck1, FakeCheck2] + monkeypatch.setattr('openshift_checks.OpenShiftCheck.subclasses', classmethod(lambda cls: checks)) + + result = plugin.run(tmp=None, task_vars=task_vars) + + assert failed(result, msg_has=['duplicate', 'duplicate_name', 'FakeCheck']) + + +def test_action_plugin_skip_non_active_checks(plugin, task_vars, monkeypatch): + checks = [fake_check(is_active=False)] + monkeypatch.setattr('openshift_checks.OpenShiftCheck.subclasses', classmethod(lambda cls: checks)) + + result = plugin.run(tmp=None, task_vars=task_vars) + + assert result['checks']['fake_check'] == dict(skipped=True, skipped_reason="Not active for this host") + assert not failed(result) + assert not changed(result) + assert not skipped(result) + + +def test_action_plugin_skip_disabled_checks(plugin, task_vars, monkeypatch): + checks = [fake_check('fake_check', is_active=True)] + monkeypatch.setattr('openshift_checks.OpenShiftCheck.subclasses', classmethod(lambda cls: checks)) + + task_vars['openshift_disable_check'] = 'fake_check' + result = plugin.run(tmp=None, task_vars=task_vars) + + assert result['checks']['fake_check'] == dict(skipped=True, skipped_reason="Disabled by user request") + assert not failed(result) + assert not changed(result) + assert not skipped(result) + + +def test_action_plugin_run_check_ok(plugin, task_vars, monkeypatch): + check_return_value = {'ok': 'test'} + check_class = fake_check(run_return=check_return_value) + monkeypatch.setattr(plugin, 'load_known_checks', lambda tmp, task_vars: {'fake_check': check_class()}) + monkeypatch.setattr('openshift_health_check.resolve_checks', lambda *args: ['fake_check']) + + result = plugin.run(tmp=None, task_vars=task_vars) + + assert result['checks']['fake_check'] == check_return_value + assert not failed(result) + assert not changed(result) + assert not skipped(result) + + +def test_action_plugin_run_check_changed(plugin, task_vars, monkeypatch): + check_return_value = {'ok': 'test'} + check_class = fake_check(run_return=check_return_value, changed=True) + monkeypatch.setattr(plugin, 'load_known_checks', lambda tmp, task_vars: {'fake_check': check_class()}) + monkeypatch.setattr('openshift_health_check.resolve_checks', lambda *args: ['fake_check']) + + result = plugin.run(tmp=None, task_vars=task_vars) + + assert result['checks']['fake_check'] == check_return_value + assert changed(result['checks']['fake_check']) + assert not failed(result) + assert changed(result) + assert not skipped(result) + + +def test_action_plugin_run_check_fail(plugin, task_vars, monkeypatch): + check_return_value = {'failed': True} + check_class = fake_check(run_return=check_return_value) + monkeypatch.setattr(plugin, 'load_known_checks', lambda tmp, task_vars: {'fake_check': check_class()}) + monkeypatch.setattr('openshift_health_check.resolve_checks', lambda *args: ['fake_check']) + + result = plugin.run(tmp=None, task_vars=task_vars) + + assert result['checks']['fake_check'] == check_return_value + assert failed(result, msg_has=['failed']) + assert not changed(result) + assert not skipped(result) + + +def test_action_plugin_run_check_exception(plugin, task_vars, monkeypatch): + exception_msg = 'fake check has an exception' + run_exception = OpenShiftCheckException(exception_msg) + check_class = fake_check(run_exception=run_exception, changed=True) + monkeypatch.setattr(plugin, 'load_known_checks', lambda tmp, task_vars: {'fake_check': check_class()}) + monkeypatch.setattr('openshift_health_check.resolve_checks', lambda *args: ['fake_check']) + + result = plugin.run(tmp=None, task_vars=task_vars) + + assert failed(result['checks']['fake_check'], msg_has=exception_msg) + assert failed(result, msg_has=['failed']) + assert changed(result['checks']['fake_check']) + assert changed(result) + assert not skipped(result) + + +def test_action_plugin_resolve_checks_exception(plugin, task_vars, monkeypatch): + monkeypatch.setattr(plugin, 'load_known_checks', lambda tmp, task_vars: {}) + + result = plugin.run(tmp=None, task_vars=task_vars) + + assert failed(result, msg_has=['unknown', 'name']) + assert not changed(result) + assert not skipped(result) + + +@pytest.mark.parametrize('names,all_checks,expected', [ + ([], [], set()), + ( + ['a', 'b'], + [ + fake_check('a'), + fake_check('b'), + ], + set(['a', 'b']), + ), + ( + ['a', 'b', '@group'], + [ + fake_check('from_group_1', ['group', 'another_group']), + fake_check('not_in_group', ['another_group']), + fake_check('from_group_2', ['preflight', 'group']), + fake_check('a'), + fake_check('b'), + ], + set(['a', 'b', 'from_group_1', 'from_group_2']), + ), +]) +def test_resolve_checks_ok(names, all_checks, expected): + assert resolve_checks(names, all_checks) == expected + + +@pytest.mark.parametrize('names,all_checks,words_in_exception', [ + ( + ['testA', 'testB'], + [], + ['check', 'name', 'testA', 'testB'], + ), + ( + ['@group'], + [], + ['tag', 'name', 'group'], + ), + ( + ['testA', 'testB', '@group'], + [], + ['check', 'name', 'testA', 'testB', 'tag', 'group'], + ), + ( + ['testA', 'testB', '@group'], + [ + fake_check('from_group_1', ['group', 'another_group']), + fake_check('not_in_group', ['another_group']), + fake_check('from_group_2', ['preflight', 'group']), + ], + ['check', 'name', 'testA', 'testB'], + ), +]) +def test_resolve_checks_failure(names, all_checks, words_in_exception): + with pytest.raises(Exception) as excinfo: + resolve_checks(names, all_checks) + for word in words_in_exception: + assert word in str(excinfo.value) diff --git a/roles/openshift_health_checker/test/aos_version_test.py b/roles/openshift_health_checker/test/aos_version_test.py new file mode 100644 index 000000000..4100f6c70 --- /dev/null +++ b/roles/openshift_health_checker/test/aos_version_test.py @@ -0,0 +1,196 @@ +import pytest +import aos_version + +from collections import namedtuple +Package = namedtuple('Package', ['name', 'version']) + +expected_pkgs = { + "spam": { + "name": "spam", + "version": "3.2.1", + "check_multi": False, + }, + "eggs": { + "name": "eggs", + "version": "3.2.1", + "check_multi": False, + }, +} + + +@pytest.mark.parametrize('pkgs,expected_pkgs_dict', [ + ( + # all found + [Package('spam', '3.2.1'), Package('eggs', '3.2.1')], + expected_pkgs, + ), + ( + # found with more specific version + [Package('spam', '3.2.1'), Package('eggs', '3.2.1.5')], + expected_pkgs, + ), + ( + [Package('ovs', '2.6'), Package('ovs', '2.4')], + { + "ovs": { + "name": "ovs", + "version": ["2.6", "2.7"], + "check_multi": False, + } + }, + ), + ( + [Package('ovs', '2.7')], + { + "ovs": { + "name": "ovs", + "version": ["2.6", "2.7"], + "check_multi": False, + } + }, + ), +]) +def test_check_precise_version_found(pkgs, expected_pkgs_dict): + aos_version._check_precise_version_found(pkgs, expected_pkgs_dict) + + +@pytest.mark.parametrize('pkgs,expect_not_found', [ + ( + [], + { + "spam": { + "name": "spam", + "version": "3.2.1", + "check_multi": False, + }, + "eggs": { + "name": "eggs", + "version": "3.2.1", + "check_multi": False, + } + }, # none found + ), + ( + [Package('spam', '3.2.1')], + { + "eggs": { + "name": "eggs", + "version": "3.2.1", + "check_multi": False, + } + }, # completely missing + ), + ( + [Package('spam', '3.2.1'), Package('eggs', '3.3.2')], + { + "eggs": { + "name": "eggs", + "version": "3.2.1", + "check_multi": False, + } + }, # not the right version + ), + ( + [Package('eggs', '1.2.3'), Package('eggs', '3.2.1.5')], + { + "spam": { + "name": "spam", + "version": "3.2.1", + "check_multi": False, + } + }, # eggs found with multiple versions + ), +]) +def test_check_precise_version_found_fail(pkgs, expect_not_found): + with pytest.raises(aos_version.PreciseVersionNotFound) as e: + aos_version._check_precise_version_found(pkgs, expected_pkgs) + assert list(expect_not_found.values()) == e.value.problem_pkgs + + +@pytest.mark.parametrize('pkgs,expected_pkgs_dict', [ + ( + [], + expected_pkgs, + ), + ( + # more precise but not strictly higher + [Package('spam', '3.2.1.9')], + expected_pkgs, + ), + ( + [Package('ovs', '2.7')], + { + "ovs": { + "name": "ovs", + "version": ["2.6", "2.7"], + "check_multi": False, + } + }, + ), +]) +def test_check_higher_version_found(pkgs, expected_pkgs_dict): + aos_version._check_higher_version_found(pkgs, expected_pkgs_dict) + + +@pytest.mark.parametrize('pkgs,expected_pkgs_dict,expect_higher', [ + ( + [Package('spam', '3.3')], + expected_pkgs, + ['spam-3.3'], # lower precision, but higher + ), + ( + [Package('spam', '3.2.1'), Package('eggs', '3.3.2')], + expected_pkgs, + ['eggs-3.3.2'], # one too high + ), + ( + [Package('eggs', '1.2.3'), Package('eggs', '3.2.1.5'), Package('eggs', '3.4')], + expected_pkgs, + ['eggs-3.4'], # multiple versions, one is higher + ), + ( + [Package('eggs', '3.2.1'), Package('eggs', '3.4'), Package('eggs', '3.3')], + expected_pkgs, + ['eggs-3.4'], # multiple versions, two are higher + ), + ( + [Package('ovs', '2.8')], + { + "ovs": { + "name": "ovs", + "version": ["2.6", "2.7"], + "check_multi": False, + } + }, + ['ovs-2.8'], + ), +]) +def test_check_higher_version_found_fail(pkgs, expected_pkgs_dict, expect_higher): + with pytest.raises(aos_version.FoundHigherVersion) as e: + aos_version._check_higher_version_found(pkgs, expected_pkgs_dict) + assert set(expect_higher) == set(e.value.problem_pkgs) + + +@pytest.mark.parametrize('pkgs', [ + [], + [Package('spam', '3.2.1')], + [Package('spam', '3.2.1'), Package('eggs', '3.2.2')], +]) +def test_check_multi_minor_release(pkgs): + aos_version._check_multi_minor_release(pkgs, expected_pkgs) + + +@pytest.mark.parametrize('pkgs,expect_to_flag_pkgs', [ + ( + [Package('spam', '3.2.1'), Package('spam', '3.3.2')], + ['spam'], + ), + ( + [Package('eggs', '1.2.3'), Package('eggs', '3.2.1.5'), Package('eggs', '3.4')], + ['eggs'], + ), +]) +def test_check_multi_minor_release_fail(pkgs, expect_to_flag_pkgs): + with pytest.raises(aos_version.FoundMultiRelease) as e: + aos_version._check_multi_minor_release(pkgs, expected_pkgs) + assert set(expect_to_flag_pkgs) == set(e.value.problem_pkgs) diff --git a/roles/openshift_health_checker/test/conftest.py b/roles/openshift_health_checker/test/conftest.py new file mode 100644 index 000000000..244a1f0fa --- /dev/null +++ b/roles/openshift_health_checker/test/conftest.py @@ -0,0 +1,12 @@ +import os +import sys + +# extend sys.path so that tests can import openshift_checks and action plugins +# from this role. +openshift_health_checker_path = os.path.dirname(os.path.dirname(__file__)) +sys.path[1:1] = [ + openshift_health_checker_path, + os.path.join(openshift_health_checker_path, 'action_plugins'), + os.path.join(openshift_health_checker_path, 'callback_plugins'), + os.path.join(openshift_health_checker_path, 'library'), +] diff --git a/roles/openshift_health_checker/test/curator_test.py b/roles/openshift_health_checker/test/curator_test.py new file mode 100644 index 000000000..62c680b74 --- /dev/null +++ b/roles/openshift_health_checker/test/curator_test.py @@ -0,0 +1,57 @@ +import pytest + +from openshift_checks.logging.curator import Curator, OpenShiftCheckException + + +plain_curator_pod = { + "metadata": { + "labels": {"component": "curator", "deploymentconfig": "logging-curator"}, + "name": "logging-curator-1", + }, + "status": { + "containerStatuses": [{"ready": True}], + "conditions": [{"status": "True", "type": "Ready"}], + "podIP": "10.10.10.10", + } +} + +not_running_curator_pod = { + "metadata": { + "labels": {"component": "curator", "deploymentconfig": "logging-curator"}, + "name": "logging-curator-2", + }, + "status": { + "containerStatuses": [{"ready": False}], + "conditions": [{"status": "False", "type": "Ready"}], + "podIP": "10.10.10.10", + } +} + + +def test_get_curator_pods(): + check = Curator() + check.get_pods_for_component = lambda *_: [plain_curator_pod] + result = check.run() + assert "failed" not in result or not result["failed"] + + +@pytest.mark.parametrize('pods, expect_error', [ + ( + [], + 'MissingComponentPods', + ), + ( + [not_running_curator_pod], + 'CuratorNotRunning', + ), + ( + [plain_curator_pod, plain_curator_pod], + 'TooManyCurators', + ), +]) +def test_get_curator_pods_fail(pods, expect_error): + check = Curator() + check.get_pods_for_component = lambda *_: pods + with pytest.raises(OpenShiftCheckException) as excinfo: + check.run() + assert excinfo.value.name == expect_error diff --git a/roles/openshift_health_checker/test/disk_availability_test.py b/roles/openshift_health_checker/test/disk_availability_test.py new file mode 100644 index 000000000..f4fd2dfed --- /dev/null +++ b/roles/openshift_health_checker/test/disk_availability_test.py @@ -0,0 +1,248 @@ +import pytest + +from openshift_checks.disk_availability import DiskAvailability, OpenShiftCheckException + + +@pytest.mark.parametrize('group_names,is_active', [ + (['masters'], True), + (['nodes'], True), + (['etcd'], True), + (['masters', 'nodes'], True), + (['masters', 'etcd'], True), + ([], False), + (['lb'], False), + (['nfs'], False), +]) +def test_is_active(group_names, is_active): + task_vars = dict( + group_names=group_names, + ) + assert DiskAvailability(None, task_vars).is_active() == is_active + + +@pytest.mark.parametrize('desc, ansible_mounts, expect_chunks', [ + ( + 'empty ansible_mounts', + [], + ['determine mount point', 'none'], + ), + ( + 'missing relevant mount paths', + [{'mount': '/mnt'}], + ['determine mount point', '/mnt'], + ), + ( + 'missing size_available', + [{'mount': '/var'}, {'mount': '/usr'}, {'mount': '/tmp'}], + ['missing', 'size_available'], + ), +]) +def test_cannot_determine_available_disk(desc, ansible_mounts, expect_chunks): + task_vars = dict( + group_names=['masters'], + ansible_mounts=ansible_mounts, + ) + + with pytest.raises(OpenShiftCheckException) as excinfo: + DiskAvailability(fake_execute_module, task_vars).run() + + for chunk in expect_chunks: + assert chunk in str(excinfo.value) + + +@pytest.mark.parametrize('group_names,configured_min,ansible_mounts', [ + ( + ['masters'], + 0, + [{ + 'mount': '/', + 'size_available': 40 * 10**9 + 1, + }], + ), + ( + ['nodes'], + 0, + [{ + 'mount': '/', + 'size_available': 15 * 10**9 + 1, + }], + ), + ( + ['etcd'], + 0, + [{ + 'mount': '/', + 'size_available': 20 * 10**9 + 1, + }], + ), + ( + ['etcd'], + 1, # configure lower threshold + [{ + 'mount': '/', + 'size_available': 1 * 10**9 + 1, # way smaller than recommended + }], + ), + ( + ['etcd'], + 0, + [{ + # not enough space on / ... + 'mount': '/', + 'size_available': 2 * 10**9, + }, { + # ... but enough on /var + 'mount': '/var', + 'size_available': 20 * 10**9 + 1, + }], + ), +]) +def test_succeeds_with_recommended_disk_space(group_names, configured_min, ansible_mounts): + task_vars = dict( + group_names=group_names, + openshift_check_min_host_disk_gb=configured_min, + ansible_mounts=ansible_mounts, + ) + + result = DiskAvailability(fake_execute_module, task_vars).run() + + assert not result.get('failed', False) + + +@pytest.mark.parametrize('name,group_names,configured_min,ansible_mounts,expect_chunks', [ + ( + 'test with no space available', + ['masters'], + 0, + [{ + 'mount': '/', + 'size_available': 1, + }], + ['0.0 GB'], + ), + ( + 'test with a higher configured required value', + ['masters'], + 100, # set a higher threshold + [{ + 'mount': '/', + 'size_available': 50 * 10**9, # would normally be enough... + }], + ['100.0 GB'], + ), + ( + 'test with 1GB available, but "0" GB space requirement', + ['nodes'], + 0, + [{ + 'mount': '/', + 'size_available': 1 * 10**9, + }], + ['1.0 GB'], + ), + ( + 'test with no space available, but "0" GB space requirement', + ['etcd'], + 0, + [{ + 'mount': '/', + 'size_available': 1, + }], + ['0.0 GB'], + ), + ( + 'test with enough space for a node, but not for a master', + ['nodes', 'masters'], + 0, + [{ + 'mount': '/', + 'size_available': 15 * 10**9 + 1, + }], + ['15.0 GB'], + ), + ( + 'test failure with enough space on "/", but not enough on "/var"', + ['etcd'], + 0, + [{ + # enough space on / ... + 'mount': '/', + 'size_available': 20 * 10**9 + 1, + }, { + # .. but not enough on /var + 'mount': '/var', + 'size_available': 0, + }], + ['0.0 GB'], + ), +], ids=lambda argval: argval[0]) +def test_fails_with_insufficient_disk_space(name, group_names, configured_min, ansible_mounts, expect_chunks): + task_vars = dict( + group_names=group_names, + openshift_check_min_host_disk_gb=configured_min, + ansible_mounts=ansible_mounts, + ) + + result = DiskAvailability(fake_execute_module, task_vars).run() + + assert result['failed'] + for chunk in 'below recommended'.split() + expect_chunks: + assert chunk in result.get('msg', '') + + +@pytest.mark.parametrize('name,group_names,context,ansible_mounts,failed,extra_words', [ + ( + 'test without enough space for master under "upgrade" context', + ['nodes', 'masters'], + "upgrade", + [{ + 'mount': '/', + 'size_available': 1 * 10**9 + 1, + 'size_total': 21 * 10**9 + 1, + }], + True, + ["1.0 GB"], + ), + ( + 'test with enough space for master under "upgrade" context', + ['nodes', 'masters'], + "upgrade", + [{ + 'mount': '/', + 'size_available': 10 * 10**9 + 1, + 'size_total': 21 * 10**9 + 1, + }], + False, + [], + ), + ( + 'test with not enough space for master, and non-upgrade context', + ['nodes', 'masters'], + "health", + [{ + 'mount': '/', + # not enough space for a master, + # "health" context should not lower requirement + 'size_available': 20 * 10**9 + 1, + }], + True, + ["20.0 GB", "below minimum"], + ), +], ids=lambda argval: argval[0]) +def test_min_required_space_changes_with_upgrade_context(name, group_names, context, ansible_mounts, failed, extra_words): + task_vars = dict( + r_openshift_health_checker_playbook_context=context, + group_names=group_names, + ansible_mounts=ansible_mounts, + ) + + check = DiskAvailability(fake_execute_module, task_vars) + result = check.run() + + assert result.get("failed", False) == failed + for word in extra_words: + assert word in result.get('msg', '') + + +def fake_execute_module(*args): + raise AssertionError('this function should not be called') diff --git a/roles/openshift_health_checker/test/docker_image_availability_test.py b/roles/openshift_health_checker/test/docker_image_availability_test.py new file mode 100644 index 000000000..8d0a53df9 --- /dev/null +++ b/roles/openshift_health_checker/test/docker_image_availability_test.py @@ -0,0 +1,274 @@ +import pytest + +from openshift_checks.docker_image_availability import DockerImageAvailability + + +@pytest.mark.parametrize('deployment_type, is_containerized, group_names, expect_active', [ + ("origin", True, [], True), + ("openshift-enterprise", True, [], True), + ("enterprise", True, [], False), + ("online", True, [], False), + ("invalid", True, [], False), + ("", True, [], False), + ("origin", False, [], False), + ("openshift-enterprise", False, [], False), + ("origin", False, ["nodes", "masters"], True), + ("openshift-enterprise", False, ["etcd"], False), +]) +def test_is_active(deployment_type, is_containerized, group_names, expect_active): + task_vars = dict( + openshift=dict(common=dict(is_containerized=is_containerized)), + openshift_deployment_type=deployment_type, + group_names=group_names, + ) + assert DockerImageAvailability(None, task_vars).is_active() == expect_active + + +@pytest.mark.parametrize("is_containerized,is_atomic", [ + (True, True), + (False, False), + (True, False), + (False, True), +]) +def test_all_images_available_locally(is_containerized, is_atomic): + def execute_module(module_name, module_args, *_): + if module_name == "yum": + return {"changed": True} + + assert module_name == "docker_image_facts" + assert 'name' in module_args + assert module_args['name'] + return { + 'images': [module_args['name']], + } + + result = DockerImageAvailability(execute_module, task_vars=dict( + openshift=dict( + common=dict( + service_type='origin', + is_containerized=is_containerized, + is_atomic=is_atomic, + ), + docker=dict(additional_registries=["docker.io"]), + ), + openshift_deployment_type='origin', + openshift_image_tag='3.4', + group_names=['nodes', 'masters'], + )).run() + + assert not result.get('failed', False) + + +@pytest.mark.parametrize("available_locally", [ + False, + True, +]) +def test_all_images_available_remotely(available_locally): + def execute_module(module_name, *_): + if module_name == 'docker_image_facts': + return {'images': [], 'failed': available_locally} + return {'changed': False} + + result = DockerImageAvailability(execute_module, task_vars=dict( + openshift=dict( + common=dict( + service_type='origin', + is_containerized=False, + is_atomic=False, + ), + docker=dict(additional_registries=["docker.io", "registry.access.redhat.com"]), + ), + openshift_deployment_type='origin', + openshift_image_tag='v3.4', + group_names=['nodes', 'masters'], + )).run() + + assert not result.get('failed', False) + + +def test_all_images_unavailable(): + def execute_module(module_name=None, *_): + if module_name == "command": + return { + 'failed': True, + } + + return { + 'changed': False, + } + + actual = DockerImageAvailability(execute_module, task_vars=dict( + openshift=dict( + common=dict( + service_type='origin', + is_containerized=False, + is_atomic=False, + ), + docker=dict(additional_registries=["docker.io"]), + ), + openshift_deployment_type="openshift-enterprise", + openshift_image_tag='latest', + group_names=['nodes', 'masters'], + )).run() + + assert actual['failed'] + assert "required Docker images are not available" in actual['msg'] + + +@pytest.mark.parametrize("message,extra_words", [ + ( + "docker image update failure", + ["docker image update failure"], + ), + ( + "No package matching 'skopeo' found available, installed or updated", + ["dependencies can be installed via `yum`"] + ), +]) +def test_skopeo_update_failure(message, extra_words): + def execute_module(module_name=None, *_): + if module_name == "yum": + return { + "failed": True, + "msg": message, + "changed": False, + } + + return {'changed': False} + + actual = DockerImageAvailability(execute_module, task_vars=dict( + openshift=dict( + common=dict( + service_type='origin', + is_containerized=False, + is_atomic=False, + ), + docker=dict(additional_registries=["unknown.io"]), + ), + openshift_deployment_type="openshift-enterprise", + openshift_image_tag='', + group_names=['nodes', 'masters'], + )).run() + + assert actual["failed"] + for word in extra_words: + assert word in actual["msg"] + + +@pytest.mark.parametrize("deployment_type,registries", [ + ("origin", ["unknown.io"]), + ("openshift-enterprise", ["registry.access.redhat.com"]), + ("openshift-enterprise", []), +]) +def test_registry_availability(deployment_type, registries): + def execute_module(module_name=None, *_): + return { + 'changed': False, + } + + actual = DockerImageAvailability(execute_module, task_vars=dict( + openshift=dict( + common=dict( + service_type='origin', + is_containerized=False, + is_atomic=False, + ), + docker=dict(additional_registries=registries), + ), + openshift_deployment_type=deployment_type, + openshift_image_tag='', + group_names=['nodes', 'masters'], + )).run() + + assert not actual.get("failed", False) + + +@pytest.mark.parametrize("deployment_type, is_containerized, groups, oreg_url, expected", [ + ( # standard set of stuff required on nodes + "origin", False, ['nodes'], None, + set([ + 'openshift/origin-pod:vtest', + 'openshift/origin-deployer:vtest', + 'openshift/origin-docker-registry:vtest', + 'openshift/origin-haproxy-router:vtest', + 'cockpit/kubernetes', # origin version of registry-console + ]) + ), + ( # set a different URL for images + "origin", False, ['nodes'], 'foo.io/openshift/origin-${component}:${version}', + set([ + 'foo.io/openshift/origin-pod:vtest', + 'foo.io/openshift/origin-deployer:vtest', + 'foo.io/openshift/origin-docker-registry:vtest', + 'foo.io/openshift/origin-haproxy-router:vtest', + 'cockpit/kubernetes', # AFAICS this is not built from the URL + ]) + ), + ( + "origin", True, ['nodes', 'masters', 'etcd'], None, + set([ + # images running on top of openshift + 'openshift/origin-pod:vtest', + 'openshift/origin-deployer:vtest', + 'openshift/origin-docker-registry:vtest', + 'openshift/origin-haproxy-router:vtest', + 'cockpit/kubernetes', + # containerized component images + 'openshift/origin:vtest', + 'openshift/node:vtest', + 'openshift/openvswitch:vtest', + 'registry.access.redhat.com/rhel7/etcd', + ]) + ), + ( # enterprise images + "openshift-enterprise", True, ['nodes'], 'foo.io/openshift3/ose-${component}:f13ac45', + set([ + 'foo.io/openshift3/ose-pod:f13ac45', + 'foo.io/openshift3/ose-deployer:f13ac45', + 'foo.io/openshift3/ose-docker-registry:f13ac45', + 'foo.io/openshift3/ose-haproxy-router:f13ac45', + # registry-console is not constructed/versioned the same as the others. + 'registry.access.redhat.com/openshift3/registry-console', + # containerized images aren't built from oreg_url + 'openshift3/node:vtest', + 'openshift3/openvswitch:vtest', + ]) + ), + ( + "openshift-enterprise", True, ['etcd', 'lb'], 'foo.io/openshift3/ose-${component}:f13ac45', + set([ + 'registry.access.redhat.com/rhel7/etcd', + # lb does not yet come in a containerized version + ]) + ), + +]) +def test_required_images(deployment_type, is_containerized, groups, oreg_url, expected): + task_vars = dict( + openshift=dict( + common=dict( + is_containerized=is_containerized, + is_atomic=False, + ), + ), + openshift_deployment_type=deployment_type, + group_names=groups, + oreg_url=oreg_url, + openshift_image_tag='vtest', + ) + + assert expected == DockerImageAvailability("DUMMY", task_vars).required_images() + + +def test_containerized_etcd(): + task_vars = dict( + openshift=dict( + common=dict( + is_containerized=True, + ), + ), + openshift_deployment_type="origin", + group_names=['etcd'], + ) + expected = set(['registry.access.redhat.com/rhel7/etcd']) + assert expected == DockerImageAvailability("DUMMY", task_vars).required_images() diff --git a/roles/openshift_health_checker/test/docker_storage_test.py b/roles/openshift_health_checker/test/docker_storage_test.py new file mode 100644 index 000000000..e0dccc062 --- /dev/null +++ b/roles/openshift_health_checker/test/docker_storage_test.py @@ -0,0 +1,305 @@ +import pytest + +from openshift_checks import OpenShiftCheckException +from openshift_checks.docker_storage import DockerStorage + + +@pytest.mark.parametrize('is_containerized, group_names, is_active', [ + (False, ["masters", "etcd"], False), + (False, ["masters", "nodes"], True), + (True, ["etcd"], True), +]) +def test_is_active(is_containerized, group_names, is_active): + task_vars = dict( + openshift=dict(common=dict(is_containerized=is_containerized)), + group_names=group_names, + ) + assert DockerStorage(None, task_vars).is_active() == is_active + + +def non_atomic_task_vars(): + return {"openshift": {"common": {"is_atomic": False}}} + + +@pytest.mark.parametrize('docker_info, failed, expect_msg', [ + ( + dict(failed=True, msg="Error connecting: Error while fetching server API version"), + True, + ["Is docker running on this host?"], + ), + ( + dict(msg="I have no info"), + True, + ["missing info"], + ), + ( + dict(info={ + "Driver": "devicemapper", + "DriverStatus": [("Pool Name", "docker-docker--pool")], + }), + False, + [], + ), + ( + dict(info={ + "Driver": "devicemapper", + "DriverStatus": [("Data loop file", "true")], + }), + True, + ["loopback devices with the Docker devicemapper storage driver"], + ), + ( + dict(info={ + "Driver": "overlay2", + "DriverStatus": [("Backing Filesystem", "xfs")], + }), + False, + [], + ), + ( + dict(info={ + "Driver": "overlay", + "DriverStatus": [("Backing Filesystem", "btrfs")], + }), + True, + ["storage is type 'btrfs'", "only supported with\n'xfs'"], + ), + ( + dict(info={ + "Driver": "overlay2", + "DriverStatus": [("Backing Filesystem", "xfs")], + "OperatingSystem": "Red Hat Enterprise Linux Server release 7.2 (Maipo)", + "KernelVersion": "3.10.0-327.22.2.el7.x86_64", + }), + True, + ["Docker reports kernel version 3.10.0-327"], + ), + ( + dict(info={ + "Driver": "overlay", + "DriverStatus": [("Backing Filesystem", "xfs")], + "OperatingSystem": "CentOS", + "KernelVersion": "3.10.0-514", + }), + False, + [], + ), + ( + dict(info={ + "Driver": "unsupported", + }), + True, + ["unsupported Docker storage driver"], + ), +]) +def test_check_storage_driver(docker_info, failed, expect_msg): + def execute_module(module_name, *_): + if module_name == "yum": + return {} + if module_name != "docker_info": + raise ValueError("not expecting module " + module_name) + return docker_info + + check = DockerStorage(execute_module, non_atomic_task_vars()) + check.check_dm_usage = lambda status: dict() # stub out for this test + check.check_overlay_usage = lambda info: dict() # stub out for this test + result = check.run() + + if failed: + assert result["failed"] + else: + assert not result.get("failed", False) + + for word in expect_msg: + assert word in result["msg"] + + +enough_space = { + "Pool Name": "docker--vg-docker--pool", + "Data Space Used": "19.92 MB", + "Data Space Total": "8.535 GB", + "Metadata Space Used": "40.96 kB", + "Metadata Space Total": "25.17 MB", +} + +not_enough_space = { + "Pool Name": "docker--vg-docker--pool", + "Data Space Used": "10 GB", + "Data Space Total": "10 GB", + "Metadata Space Used": "42 kB", + "Metadata Space Total": "43 kB", +} + + +@pytest.mark.parametrize('task_vars, driver_status, vg_free, success, expect_msg', [ + ( + {"max_thinpool_data_usage_percent": "not a float"}, + enough_space, + "12g", + False, + ["is not a percentage"], + ), + ( + {}, + {}, # empty values from driver status + "bogus", # also does not parse as bytes + False, + ["Could not interpret", "as bytes"], + ), + ( + {}, + enough_space, + "12.00g", + True, + [], + ), + ( + {}, + not_enough_space, + "0.00", + False, + ["data usage", "metadata usage", "higher than threshold"], + ), +]) +def test_dm_usage(task_vars, driver_status, vg_free, success, expect_msg): + check = DockerStorage(None, task_vars) + check.get_vg_free = lambda pool: vg_free + result = check.check_dm_usage(driver_status) + result_success = not result.get("failed") + + assert result_success is success + for msg in expect_msg: + assert msg in result["msg"] + + +@pytest.mark.parametrize('pool, command_returns, raises, returns', [ + ( + "foo-bar", + { # vgs missing + "msg": "[Errno 2] No such file or directory", + "failed": True, + "cmd": "/sbin/vgs", + "rc": 2, + }, + "Failed to run /sbin/vgs", + None, + ), + ( + "foo", # no hyphen in name - should not happen + {}, + "name does not have the expected format", + None, + ), + ( + "foo-bar", + dict(stdout=" 4.00g\n"), + None, + "4.00g", + ), + ( + "foo-bar", + dict(stdout="\n"), # no matching VG + "vgs did not find this VG", + None, + ) +]) +def test_vg_free(pool, command_returns, raises, returns): + def execute_module(module_name, *_): + if module_name != "command": + raise ValueError("not expecting module " + module_name) + return command_returns + + check = DockerStorage(execute_module) + if raises: + with pytest.raises(OpenShiftCheckException) as err: + check.get_vg_free(pool) + assert raises in str(err.value) + else: + ret = check.get_vg_free(pool) + assert ret == returns + + +@pytest.mark.parametrize('string, expect_bytes', [ + ("12", 12.0), + ("12 k", 12.0 * 1024), + ("42.42 MB", 42.42 * 1024**2), + ("12g", 12.0 * 1024**3), +]) +def test_convert_to_bytes(string, expect_bytes): + got = DockerStorage.convert_to_bytes(string) + assert got == expect_bytes + + +@pytest.mark.parametrize('string', [ + "bork", + "42 Qs", +]) +def test_convert_to_bytes_error(string): + with pytest.raises(ValueError) as err: + DockerStorage.convert_to_bytes(string) + assert "Cannot convert" in str(err.value) + assert string in str(err.value) + + +ansible_mounts_enough = [{ + 'mount': '/var/lib/docker', + 'size_available': 50 * 10**9, + 'size_total': 50 * 10**9, +}] +ansible_mounts_not_enough = [{ + 'mount': '/var/lib/docker', + 'size_available': 0, + 'size_total': 50 * 10**9, +}] +ansible_mounts_missing_fields = [dict(mount='/var/lib/docker')] +ansible_mounts_zero_size = [{ + 'mount': '/var/lib/docker', + 'size_available': 0, + 'size_total': 0, +}] + + +@pytest.mark.parametrize('ansible_mounts, threshold, expect_fail, expect_msg', [ + ( + ansible_mounts_enough, + None, + False, + [], + ), + ( + ansible_mounts_not_enough, + None, + True, + ["usage percentage", "higher than threshold"], + ), + ( + ansible_mounts_not_enough, + "bogus percent", + True, + ["is not a percentage"], + ), + ( + ansible_mounts_missing_fields, + None, + True, + ["Ansible bug"], + ), + ( + ansible_mounts_zero_size, + None, + True, + ["Ansible bug"], + ), +]) +def test_overlay_usage(ansible_mounts, threshold, expect_fail, expect_msg): + task_vars = non_atomic_task_vars() + task_vars["ansible_mounts"] = ansible_mounts + if threshold is not None: + task_vars["max_overlay_usage_percent"] = threshold + check = DockerStorage(None, task_vars) + docker_info = dict(DockerRootDir="/var/lib/docker", Driver="overlay") + result = check.check_overlay_usage(docker_info) + + assert expect_fail == bool(result.get("failed")) + for msg in expect_msg: + assert msg in result["msg"] diff --git a/roles/openshift_health_checker/test/elasticsearch_test.py b/roles/openshift_health_checker/test/elasticsearch_test.py new file mode 100644 index 000000000..09bacd9ac --- /dev/null +++ b/roles/openshift_health_checker/test/elasticsearch_test.py @@ -0,0 +1,203 @@ +import pytest +import json + +from openshift_checks.logging.elasticsearch import Elasticsearch, OpenShiftCheckExceptionList + + +task_vars_config_base = dict(openshift=dict(common=dict(config_base='/etc/origin'))) + + +def canned_elasticsearch(task_vars=None, exec_oc=None): + """Create an Elasticsearch check object with stubbed exec_oc method""" + check = Elasticsearch(None, task_vars or {}) + if exec_oc: + check.exec_oc = exec_oc + return check + + +def assert_error_in_list(expect_err, errorlist): + assert any(err.name == expect_err for err in errorlist), "{} in {}".format(str(expect_err), str(errorlist)) + + +def pods_by_name(pods): + return {pod['metadata']['name']: pod for pod in pods} + + +plain_es_pod = { + "metadata": { + "labels": {"component": "es", "deploymentconfig": "logging-es"}, + "name": "logging-es", + }, + "spec": {}, + "status": { + "conditions": [{"status": "True", "type": "Ready"}], + "containerStatuses": [{"ready": True}], + "podIP": "10.10.10.10", + }, + "_test_master_name_str": "name logging-es", +} + +split_es_pod = { + "metadata": { + "labels": {"component": "es", "deploymentconfig": "logging-es-2"}, + "name": "logging-es-2", + }, + "spec": {}, + "status": { + "conditions": [{"status": "True", "type": "Ready"}], + "containerStatuses": [{"ready": True}], + "podIP": "10.10.10.10", + }, + "_test_master_name_str": "name logging-es-2", +} + +unready_es_pod = { + "metadata": { + "labels": {"component": "es", "deploymentconfig": "logging-es-3"}, + "name": "logging-es-3", + }, + "spec": {}, + "status": { + "conditions": [{"status": "False", "type": "Ready"}], + "containerStatuses": [{"ready": False}], + "podIP": "10.10.10.10", + }, + "_test_master_name_str": "BAD_NAME_RESPONSE", +} + + +def test_check_elasticsearch(): + with pytest.raises(OpenShiftCheckExceptionList) as excinfo: + canned_elasticsearch().check_elasticsearch([]) + assert_error_in_list('NoRunningPods', excinfo.value) + + # canned oc responses to match so all the checks pass + def exec_oc(cmd, args): + if '_cat/master' in cmd: + return 'name logging-es' + elif '/_nodes' in cmd: + return json.dumps(es_node_list) + elif '_cluster/health' in cmd: + return '{"status": "green"}' + elif ' df ' in cmd: + return 'IUse% Use%\n 3% 4%\n' + else: + raise Exception(cmd) + + check = canned_elasticsearch({}, exec_oc) + check.get_pods_for_component = lambda *_: [plain_es_pod] + assert {} == check.run() + + +def test_check_running_es_pods(): + pods, errors = Elasticsearch().running_elasticsearch_pods([plain_es_pod, unready_es_pod]) + assert plain_es_pod in pods + assert_error_in_list('PodNotRunning', errors) + + +def test_check_elasticsearch_masters(): + pods = [plain_es_pod] + check = canned_elasticsearch(task_vars_config_base, lambda *_: plain_es_pod['_test_master_name_str']) + assert not check.check_elasticsearch_masters(pods_by_name(pods)) + + +@pytest.mark.parametrize('pods, expect_error', [ + ( + [], + 'NoMasterFound', + ), + ( + [unready_es_pod], + 'NoMasterName', + ), + ( + [plain_es_pod, split_es_pod], + 'SplitBrainMasters', + ), +]) +def test_check_elasticsearch_masters_error(pods, expect_error): + test_pods = list(pods) + check = canned_elasticsearch(task_vars_config_base, lambda *_: test_pods.pop(0)['_test_master_name_str']) + assert_error_in_list(expect_error, check.check_elasticsearch_masters(pods_by_name(pods))) + + +es_node_list = { + 'nodes': { + 'random-es-name': { + 'host': 'logging-es', + }}} + + +def test_check_elasticsearch_node_list(): + check = canned_elasticsearch(task_vars_config_base, lambda *_: json.dumps(es_node_list)) + assert not check.check_elasticsearch_node_list(pods_by_name([plain_es_pod])) + + +@pytest.mark.parametrize('pods, node_list, expect_error', [ + ( + [], + {}, + 'MissingComponentPods', + ), + ( + [plain_es_pod], + {}, # empty list of nodes triggers KeyError + 'MissingNodeList', + ), + ( + [split_es_pod], + es_node_list, + 'EsPodNodeMismatch', + ), +]) +def test_check_elasticsearch_node_list_errors(pods, node_list, expect_error): + check = canned_elasticsearch(task_vars_config_base, lambda cmd, args: json.dumps(node_list)) + assert_error_in_list(expect_error, check.check_elasticsearch_node_list(pods_by_name(pods))) + + +def test_check_elasticsearch_cluster_health(): + test_health_data = [{"status": "green"}] + check = canned_elasticsearch(exec_oc=lambda *_: json.dumps(test_health_data.pop(0))) + assert not check.check_es_cluster_health(pods_by_name([plain_es_pod])) + + +@pytest.mark.parametrize('pods, health_data, expect_error', [ + ( + [plain_es_pod], + [{"no-status": "should bomb"}], + 'BadEsResponse', + ), + ( + [plain_es_pod, split_es_pod], + [{"status": "green"}, {"status": "red"}], + 'EsClusterHealthRed', + ), +]) +def test_check_elasticsearch_cluster_health_errors(pods, health_data, expect_error): + test_health_data = list(health_data) + check = canned_elasticsearch(exec_oc=lambda *_: json.dumps(test_health_data.pop(0))) + assert_error_in_list(expect_error, check.check_es_cluster_health(pods_by_name(pods))) + + +def test_check_elasticsearch_diskspace(): + check = canned_elasticsearch(exec_oc=lambda *_: 'IUse% Use%\n 3% 4%\n') + assert not check.check_elasticsearch_diskspace(pods_by_name([plain_es_pod])) + + +@pytest.mark.parametrize('disk_data, expect_error', [ + ( + 'df: /elasticsearch/persistent: No such file or directory\n', + 'BadDfResponse', + ), + ( + 'IUse% Use%\n 95% 40%\n', + 'InodeUsageTooHigh', + ), + ( + 'IUse% Use%\n 3% 94%\n', + 'DiskUsageTooHigh', + ), +]) +def test_check_elasticsearch_diskspace_errors(disk_data, expect_error): + check = canned_elasticsearch(exec_oc=lambda *_: disk_data) + assert_error_in_list(expect_error, check.check_elasticsearch_diskspace(pods_by_name([plain_es_pod]))) diff --git a/roles/openshift_health_checker/test/etcd_imagedata_size_test.py b/roles/openshift_health_checker/test/etcd_imagedata_size_test.py new file mode 100644 index 000000000..d3aae98f2 --- /dev/null +++ b/roles/openshift_health_checker/test/etcd_imagedata_size_test.py @@ -0,0 +1,329 @@ +import pytest + +from collections import namedtuple +from openshift_checks.etcd_imagedata_size import EtcdImageDataSize +from openshift_checks import OpenShiftCheckException +from etcdkeysize import check_etcd_key_size + + +def fake_etcd_client(root): + fake_nodes = dict() + fake_etcd_node(root, fake_nodes) + + clientclass = namedtuple("client", ["read"]) + return clientclass(lambda key, recursive: fake_etcd_result(fake_nodes[key])) + + +def fake_etcd_result(fake_node): + resultclass = namedtuple("result", ["leaves"]) + if not fake_node.dir: + return resultclass([fake_node]) + + return resultclass(fake_node.leaves) + + +def fake_etcd_node(node, visited): + min_req_fields = ["dir", "key"] + fields = list(node) + leaves = [] + + if node["dir"] and node.get("leaves"): + for leaf in node["leaves"]: + leaves.append(fake_etcd_node(leaf, visited)) + + if len(set(min_req_fields) - set(fields)) > 0: + raise ValueError("fake etcd nodes require at least {} fields.".format(min_req_fields)) + + if node.get("leaves"): + node["leaves"] = leaves + + nodeclass = namedtuple("node", fields) + nodeinst = nodeclass(**node) + visited[nodeinst.key] = nodeinst + + return nodeinst + + +@pytest.mark.parametrize('ansible_mounts,extra_words', [ + ([], ['none']), # empty ansible_mounts + ([{'mount': '/mnt'}], ['/mnt']), # missing relevant mount paths +]) +def test_cannot_determine_available_mountpath(ansible_mounts, extra_words): + task_vars = dict( + ansible_mounts=ansible_mounts, + ) + check = EtcdImageDataSize(fake_execute_module, task_vars) + + with pytest.raises(OpenShiftCheckException) as excinfo: + check.run() + + for word in ['Unable to determine mount point'] + extra_words: + assert word in str(excinfo.value) + + +@pytest.mark.parametrize('ansible_mounts,tree,size_limit,should_fail,extra_words', [ + ( + # test that default image size limit evals to 1/2 * (total size in use) + [{ + 'mount': '/', + 'size_available': 40 * 10**9, + 'size_total': 80 * 10**9, + }], + {"dir": False, "key": "/", "value": "1234"}, + None, + False, + [], + ), + ( + [{ + 'mount': '/', + 'size_available': 40 * 10**9, + 'size_total': 48 * 10**9, + }], + {"dir": False, "key": "/", "value": "1234"}, + None, + False, + [], + ), + ( + # set max size limit for image data to be below total node value + # total node value is defined as the sum of the value field + # from every node + [{ + 'mount': '/', + 'size_available': 40 * 10**9, + 'size_total': 48 * 10**9, + }], + {"dir": False, "key": "/", "value": "12345678"}, + 7, + True, + ["exceeds the maximum recommended limit", "0.00 GB"], + ), + ( + [{ + 'mount': '/', + 'size_available': 48 * 10**9 - 1, + 'size_total': 48 * 10**9, + }], + {"dir": False, "key": "/", "value": "1234"}, + None, + True, + ["exceeds the maximum recommended limit", "0.00 GB"], + ) +]) +def test_check_etcd_key_size_calculates_correct_limit(ansible_mounts, tree, size_limit, should_fail, extra_words): + def execute_module(module_name, module_args, *_): + if module_name != "etcdkeysize": + return { + "changed": False, + } + + client = fake_etcd_client(tree) + s, limit_exceeded = check_etcd_key_size(client, tree["key"], module_args["size_limit_bytes"]) + + return {"size_limit_exceeded": limit_exceeded} + + task_vars = dict( + etcd_max_image_data_size_bytes=size_limit, + ansible_mounts=ansible_mounts, + openshift=dict( + master=dict(etcd_hosts=["localhost"]), + common=dict(config_base="/var/lib/origin") + ) + ) + if size_limit is None: + task_vars.pop("etcd_max_image_data_size_bytes") + + check = EtcdImageDataSize(execute_module, task_vars).run() + + if should_fail: + assert check["failed"] + + for word in extra_words: + assert word in check["msg"] + else: + assert not check.get("failed", False) + + +@pytest.mark.parametrize('ansible_mounts,tree,root_path,expected_size,extra_words', [ + ( + [{ + 'mount': '/', + 'size_available': 40 * 10**9, + 'size_total': 80 * 10**9, + }], + # test recursive size check on tree with height > 1 + { + "dir": True, + "key": "/", + "leaves": [ + {"dir": False, "key": "/foo1", "value": "1234"}, + {"dir": False, "key": "/foo2", "value": "1234"}, + {"dir": False, "key": "/foo3", "value": "1234"}, + {"dir": False, "key": "/foo4", "value": "1234"}, + { + "dir": True, + "key": "/foo5", + "leaves": [ + {"dir": False, "key": "/foo/bar1", "value": "56789"}, + {"dir": False, "key": "/foo/bar2", "value": "56789"}, + {"dir": False, "key": "/foo/bar3", "value": "56789"}, + { + "dir": True, + "key": "/foo/bar4", + "leaves": [ + {"dir": False, "key": "/foo/bar/baz1", "value": "123"}, + {"dir": False, "key": "/foo/bar/baz2", "value": "123"}, + ] + }, + ] + }, + ] + }, + "/", + 37, + [], + ), + ( + [{ + 'mount': '/', + 'size_available': 40 * 10**9, + 'size_total': 80 * 10**9, + }], + # test correct sub-tree size calculation + { + "dir": True, + "key": "/", + "leaves": [ + {"dir": False, "key": "/foo1", "value": "1234"}, + {"dir": False, "key": "/foo2", "value": "1234"}, + {"dir": False, "key": "/foo3", "value": "1234"}, + {"dir": False, "key": "/foo4", "value": "1234"}, + { + "dir": True, + "key": "/foo5", + "leaves": [ + {"dir": False, "key": "/foo/bar1", "value": "56789"}, + {"dir": False, "key": "/foo/bar2", "value": "56789"}, + {"dir": False, "key": "/foo/bar3", "value": "56789"}, + { + "dir": True, + "key": "/foo/bar4", + "leaves": [ + {"dir": False, "key": "/foo/bar/baz1", "value": "123"}, + {"dir": False, "key": "/foo/bar/baz2", "value": "123"}, + ] + }, + ] + }, + ] + }, + "/foo5", + 21, + [], + ), + ( + [{ + 'mount': '/', + 'size_available': 40 * 10**9, + 'size_total': 80 * 10**9, + }], + # test that a non-existing key is handled correctly + { + "dir": False, + "key": "/", + "value": "1234", + }, + "/missing", + 0, + [], + ), + ( + [{ + 'mount': '/', + 'size_available': 40 * 10**9, + 'size_total': 80 * 10**9, + }], + # test etcd cycle handling + { + "dir": True, + "key": "/", + "leaves": [ + {"dir": False, "key": "/foo1", "value": "1234"}, + {"dir": False, "key": "/foo2", "value": "1234"}, + {"dir": False, "key": "/foo3", "value": "1234"}, + {"dir": False, "key": "/foo4", "value": "1234"}, + { + "dir": True, + "key": "/", + "leaves": [ + {"dir": False, "key": "/foo1", "value": "1"}, + ], + }, + ] + }, + "/", + 16, + [], + ), +]) +def test_etcd_key_size_check_calculates_correct_size(ansible_mounts, tree, root_path, expected_size, extra_words): + def execute_module(module_name, module_args, *_): + if module_name != "etcdkeysize": + return { + "changed": False, + } + + client = fake_etcd_client(tree) + size, limit_exceeded = check_etcd_key_size(client, root_path, module_args["size_limit_bytes"]) + + assert size == expected_size + return { + "size_limit_exceeded": limit_exceeded, + } + + task_vars = dict( + ansible_mounts=ansible_mounts, + openshift=dict( + master=dict(etcd_hosts=["localhost"]), + common=dict(config_base="/var/lib/origin") + ) + ) + + check = EtcdImageDataSize(execute_module, task_vars).run() + assert not check.get("failed", False) + + +def test_etcdkeysize_module_failure(): + def execute_module(module_name, *_): + if module_name != "etcdkeysize": + return { + "changed": False, + } + + return { + "rc": 1, + "module_stderr": "failure", + } + + task_vars = dict( + ansible_mounts=[{ + 'mount': '/', + 'size_available': 40 * 10**9, + 'size_total': 80 * 10**9, + }], + openshift=dict( + master=dict(etcd_hosts=["localhost"]), + common=dict(config_base="/var/lib/origin") + ) + ) + + check = EtcdImageDataSize(execute_module, task_vars).run() + + assert check["failed"] + for word in "Failed to retrieve stats": + assert word in check["msg"] + + +def fake_execute_module(*args): + raise AssertionError('this function should not be called') diff --git a/roles/openshift_health_checker/test/etcd_traffic_test.py b/roles/openshift_health_checker/test/etcd_traffic_test.py new file mode 100644 index 000000000..fae3e578d --- /dev/null +++ b/roles/openshift_health_checker/test/etcd_traffic_test.py @@ -0,0 +1,72 @@ +import pytest + +from openshift_checks.etcd_traffic import EtcdTraffic + + +@pytest.mark.parametrize('group_names,version,is_active', [ + (['masters'], "3.5", False), + (['masters'], "3.6", False), + (['nodes'], "3.4", False), + (['etcd'], "3.4", True), + (['etcd'], "1.5", True), + (['etcd'], "3.1", False), + (['masters', 'nodes'], "3.5", False), + (['masters', 'etcd'], "3.5", True), + ([], "3.4", False), +]) +def test_is_active(group_names, version, is_active): + task_vars = dict( + group_names=group_names, + openshift_image_tag=version, + ) + assert EtcdTraffic(task_vars=task_vars).is_active() == is_active + + +@pytest.mark.parametrize('group_names,matched,failed,extra_words', [ + (["masters"], True, True, ["Higher than normal", "traffic"]), + (["masters", "etcd"], False, False, []), + (["etcd"], False, False, []), +]) +def test_log_matches_high_traffic_msg(group_names, matched, failed, extra_words): + def execute_module(module_name, *_): + return { + "matched": matched, + "failed": failed, + } + + task_vars = dict( + group_names=group_names, + openshift=dict( + common=dict(service_type="origin", is_containerized=False), + ) + ) + + result = EtcdTraffic(execute_module, task_vars).run() + + for word in extra_words: + assert word in result.get("msg", "") + + assert result.get("failed", False) == failed + + +@pytest.mark.parametrize('is_containerized,expected_unit_value', [ + (False, "etcd"), + (True, "etcd_container"), +]) +def test_systemd_unit_matches_deployment_type(is_containerized, expected_unit_value): + task_vars = dict( + openshift=dict( + common=dict(is_containerized=is_containerized), + ) + ) + + def execute_module(module_name, args, *_): + assert module_name == "search_journalctl" + matchers = args["log_matchers"] + + for matcher in matchers: + assert matcher["unit"] == expected_unit_value + + return {"failed": False} + + EtcdTraffic(execute_module, task_vars).run() diff --git a/roles/openshift_health_checker/test/etcd_volume_test.py b/roles/openshift_health_checker/test/etcd_volume_test.py new file mode 100644 index 000000000..077cea3ea --- /dev/null +++ b/roles/openshift_health_checker/test/etcd_volume_test.py @@ -0,0 +1,147 @@ +import pytest + +from openshift_checks.etcd_volume import EtcdVolume +from openshift_checks import OpenShiftCheckException + + +@pytest.mark.parametrize('ansible_mounts,extra_words', [ + ([], ['none']), # empty ansible_mounts + ([{'mount': '/mnt'}], ['/mnt']), # missing relevant mount paths +]) +def test_cannot_determine_available_disk(ansible_mounts, extra_words): + task_vars = dict( + ansible_mounts=ansible_mounts, + ) + + with pytest.raises(OpenShiftCheckException) as excinfo: + EtcdVolume(fake_execute_module, task_vars).run() + + for word in ['Unable to determine mount point'] + extra_words: + assert word in str(excinfo.value) + + +@pytest.mark.parametrize('size_limit,ansible_mounts', [ + ( + # if no size limit is specified, expect max usage + # limit to default to 90% of size_total + None, + [{ + 'mount': '/', + 'size_available': 40 * 10**9, + 'size_total': 80 * 10**9 + }], + ), + ( + 1, + [{ + 'mount': '/', + 'size_available': 30 * 10**9, + 'size_total': 30 * 10**9, + }], + ), + ( + 20000000000, + [{ + 'mount': '/', + 'size_available': 20 * 10**9, + 'size_total': 40 * 10**9, + }], + ), + ( + 5000000000, + [{ + # not enough space on / ... + 'mount': '/', + 'size_available': 0, + 'size_total': 0, + }, { + # not enough space on /var/lib ... + 'mount': '/var/lib', + 'size_available': 2 * 10**9, + 'size_total': 21 * 10**9, + }, { + # ... but enough on /var/lib/etcd + 'mount': '/var/lib/etcd', + 'size_available': 36 * 10**9, + 'size_total': 40 * 10**9 + }], + ) +]) +def test_succeeds_with_recommended_disk_space(size_limit, ansible_mounts): + task_vars = dict( + etcd_device_usage_threshold_percent=size_limit, + ansible_mounts=ansible_mounts, + ) + + if task_vars["etcd_device_usage_threshold_percent"] is None: + task_vars.pop("etcd_device_usage_threshold_percent") + + result = EtcdVolume(fake_execute_module, task_vars).run() + + assert not result.get('failed', False) + + +@pytest.mark.parametrize('size_limit_percent,ansible_mounts,extra_words', [ + ( + # if no size limit is specified, expect max usage + # limit to default to 90% of size_total + None, + [{ + 'mount': '/', + 'size_available': 1 * 10**9, + 'size_total': 100 * 10**9, + }], + ['99.0%'], + ), + ( + 70.0, + [{ + 'mount': '/', + 'size_available': 1 * 10**6, + 'size_total': 5 * 10**9, + }], + ['100.0%'], + ), + ( + 40.0, + [{ + 'mount': '/', + 'size_available': 2 * 10**9, + 'size_total': 6 * 10**9, + }], + ['66.7%'], + ), + ( + None, + [{ + # enough space on /var ... + 'mount': '/var', + 'size_available': 20 * 10**9, + 'size_total': 20 * 10**9, + }, { + # .. but not enough on /var/lib + 'mount': '/var/lib', + 'size_available': 1 * 10**9, + 'size_total': 20 * 10**9, + }], + ['95.0%'], + ), +]) +def test_fails_with_insufficient_disk_space(size_limit_percent, ansible_mounts, extra_words): + task_vars = dict( + etcd_device_usage_threshold_percent=size_limit_percent, + ansible_mounts=ansible_mounts, + ) + + if task_vars["etcd_device_usage_threshold_percent"] is None: + task_vars.pop("etcd_device_usage_threshold_percent") + + result = EtcdVolume(fake_execute_module, task_vars).run() + + assert result['failed'] + for word in extra_words: + assert word in result['msg'] + + +def fake_execute_module(*args): + raise AssertionError('this function should not be called') diff --git a/roles/openshift_health_checker/test/fluentd_config_test.py b/roles/openshift_health_checker/test/fluentd_config_test.py new file mode 100644 index 000000000..10db253bc --- /dev/null +++ b/roles/openshift_health_checker/test/fluentd_config_test.py @@ -0,0 +1,348 @@ +import pytest + +from openshift_checks.logging.fluentd_config import FluentdConfig, OpenShiftCheckException + + +def canned_fluentd_pod(containers): + return { + "metadata": { + "labels": {"component": "fluentd", "deploymentconfig": "logging-fluentd"}, + "name": "logging-fluentd-1", + }, + "spec": { + "host": "node1", + "nodeName": "node1", + "containers": containers, + }, + "status": { + "phase": "Running", + "containerStatuses": [{"ready": True}], + "conditions": [{"status": "True", "type": "Ready"}], + } + } + + +fluentd_pod = { + "metadata": { + "labels": {"component": "fluentd", "deploymentconfig": "logging-fluentd"}, + "name": "logging-fluentd-1", + }, + "spec": { + "host": "node1", + "nodeName": "node1", + "containers": [ + { + "name": "container1", + "env": [ + { + "name": "USE_JOURNAL", + "value": "true", + } + ], + } + ], + }, + "status": { + "phase": "Running", + "containerStatuses": [{"ready": True}], + "conditions": [{"status": "True", "type": "Ready"}], + } +} + +not_running_fluentd_pod = { + "metadata": { + "labels": {"component": "fluentd", "deploymentconfig": "logging-fluentd"}, + "name": "logging-fluentd-2", + }, + "status": { + "phase": "Unknown", + "containerStatuses": [{"ready": True}, {"ready": False}], + "conditions": [{"status": "True", "type": "Ready"}], + } +} + + +@pytest.mark.parametrize('name, use_journald, logging_driver, extra_words', [ + ( + 'test success with use_journald=false, and docker config set to use "json-file"', + False, + "json-file", + [], + ), +], ids=lambda argvals: argvals[0]) +def test_check_logging_config_non_master(name, use_journald, logging_driver, extra_words): + def execute_module(module_name, args): + if module_name == "docker_info": + return { + "info": { + "LoggingDriver": logging_driver, + } + } + + return {} + + task_vars = dict( + group_names=["nodes", "etcd"], + openshift_logging_fluentd_use_journal=use_journald, + openshift=dict( + common=dict(config_base=""), + ), + ) + + check = FluentdConfig(execute_module, task_vars) + check.execute_module = execute_module + error = check.check_logging_config() + + assert error is None + + +@pytest.mark.parametrize('name, use_journald, logging_driver, words', [ + ( + 'test failure with use_journald=false, but docker config set to use "journald"', + False, + "journald", + ['json log files', 'has been set to use "journald"'], + ), + ( + 'test failure with use_journald=false, but docker config set to use an "unsupported" driver', + False, + "unsupported", + ["json log files", 'has been set to use "unsupported"'], + ), + ( + 'test failure with use_journald=true, but docker config set to use "json-file"', + True, + "json-file", + ['logs from "journald"', 'has been set to use "json-file"'], + ), +], ids=lambda argvals: argvals[0]) +def test_check_logging_config_non_master_failed(name, use_journald, logging_driver, words): + def execute_module(module_name, args): + if module_name == "docker_info": + return { + "info": { + "LoggingDriver": logging_driver, + } + } + + return {} + + task_vars = dict( + group_names=["nodes", "etcd"], + openshift_logging_fluentd_use_journal=use_journald, + openshift=dict( + common=dict(config_base=""), + ), + ) + + check = FluentdConfig(execute_module, task_vars) + check.execute_module = execute_module + error = check.check_logging_config() + + assert error is not None + for word in words: + assert word in error + + +@pytest.mark.parametrize('name, pods, logging_driver, extra_words', [ + # use_journald returns false (not using journald), but check succeeds + # since docker is set to use json-file + ( + 'test success with use_journald=false, and docker config set to use default driver "json-file"', + [canned_fluentd_pod( + [ + { + "name": "container1", + "env": [{ + "name": "USE_JOURNAL", + "value": "false", + }], + }, + ] + )], + "json-file", + [], + ), + ( + 'test success with USE_JOURNAL env var missing and docker config set to use default driver "json-file"', + [canned_fluentd_pod( + [ + { + "name": "container1", + "env": [{ + "name": "RANDOM", + "value": "value", + }], + }, + ] + )], + "json-file", + [], + ), +], ids=lambda argvals: argvals[0]) +def test_check_logging_config_master(name, pods, logging_driver, extra_words): + def execute_module(module_name, args): + if module_name == "docker_info": + return { + "info": { + "LoggingDriver": logging_driver, + } + } + + return {} + + task_vars = dict( + group_names=["masters"], + openshift=dict( + common=dict(config_base=""), + ), + ) + + check = FluentdConfig(execute_module, task_vars) + check.execute_module = execute_module + check.get_pods_for_component = lambda _: pods + error = check.check_logging_config() + + assert error is None + + +@pytest.mark.parametrize('name, pods, logging_driver, words', [ + ( + 'test failure with use_journald=false, but docker config set to use "journald"', + [canned_fluentd_pod( + [ + { + "name": "container1", + "env": [{ + "name": "USE_JOURNAL", + "value": "false", + }], + }, + ] + )], + "journald", + ['json log files', 'has been set to use "journald"'], + ), + ( + 'test failure with use_journald=true, but docker config set to use "json-file"', + [fluentd_pod], + "json-file", + ['logs from "journald"', 'has been set to use "json-file"'], + ), + ( + 'test failure with use_journald=false, but docker set to use an "unsupported" driver', + [canned_fluentd_pod( + [ + { + "name": "container1", + "env": [{ + "name": "USE_JOURNAL", + "value": "false", + }], + }, + ] + )], + "unsupported", + ["json log files", 'has been set to use "unsupported"'], + ), + ( + 'test failure with USE_JOURNAL env var missing and docker config set to use "journald"', + [canned_fluentd_pod( + [ + { + "name": "container1", + "env": [{ + "name": "RANDOM", + "value": "value", + }], + }, + ] + )], + "journald", + ["configuration is set to", "json log files"], + ), +], ids=lambda argvals: argvals[0]) +def test_check_logging_config_master_failed(name, pods, logging_driver, words): + def execute_module(module_name, args): + if module_name == "docker_info": + return { + "info": { + "LoggingDriver": logging_driver, + } + } + + return {} + + task_vars = dict( + group_names=["masters"], + openshift=dict( + common=dict(config_base=""), + ), + ) + + check = FluentdConfig(execute_module, task_vars) + check.execute_module = execute_module + check.get_pods_for_component = lambda _: pods + error = check.check_logging_config() + + assert error is not None + for word in words: + assert word in error + + +@pytest.mark.parametrize('name, pods, response, logging_driver, extra_words', [ + ( + 'test OpenShiftCheckException with no running containers', + [canned_fluentd_pod([])], + { + "failed": True, + "result": "unexpected", + }, + "json-file", + ['no running containers'], + ), + ( + 'test OpenShiftCheckException one container and no env vars set', + [canned_fluentd_pod( + [ + { + "name": "container1", + "env": [], + }, + ] + )], + { + "failed": True, + "result": "unexpected", + }, + "json-file", + ['no environment variables'], + ), +], ids=lambda argvals: argvals[0]) +def test_check_logging_config_master_fails_on_unscheduled_deployment(name, pods, response, logging_driver, extra_words): + def execute_module(module_name, args): + if module_name == "docker_info": + return { + "info": { + "LoggingDriver": logging_driver, + } + } + + return {} + + task_vars = dict( + group_names=["masters"], + openshift=dict( + common=dict(config_base=""), + ), + ) + + check = FluentdConfig(execute_module, task_vars) + check.get_pods_for_component = lambda _: pods + + with pytest.raises(OpenShiftCheckException) as error: + check.check_logging_config() + + assert error is not None + for word in extra_words: + assert word in str(error) diff --git a/roles/openshift_health_checker/test/fluentd_test.py b/roles/openshift_health_checker/test/fluentd_test.py new file mode 100644 index 000000000..e7bf9818b --- /dev/null +++ b/roles/openshift_health_checker/test/fluentd_test.py @@ -0,0 +1,112 @@ +import pytest +import json + +from openshift_checks.logging.fluentd import Fluentd, OpenShiftCheckExceptionList, OpenShiftCheckException + + +def assert_error_in_list(expect_err, errorlist): + assert any(err.name == expect_err for err in errorlist), "{} in {}".format(str(expect_err), str(errorlist)) + + +fluentd_pod_node1 = { + "metadata": { + "labels": {"component": "fluentd", "deploymentconfig": "logging-fluentd"}, + "name": "logging-fluentd-1", + }, + "spec": {"host": "node1", "nodeName": "node1"}, + "status": { + "containerStatuses": [{"ready": True}], + "conditions": [{"status": "True", "type": "Ready"}], + } +} +fluentd_pod_node2_down = { + "metadata": { + "labels": {"component": "fluentd", "deploymentconfig": "logging-fluentd"}, + "name": "logging-fluentd-2", + }, + "spec": {"host": "node2", "nodeName": "node2"}, + "status": { + "containerStatuses": [{"ready": False}], + "conditions": [{"status": "False", "type": "Ready"}], + } +} +fluentd_node1 = { + "metadata": { + "labels": {"logging-infra-fluentd": "true", "kubernetes.io/hostname": "node1"}, + "name": "node1", + }, + "status": {"addresses": [{"type": "InternalIP", "address": "10.10.1.1"}]}, +} +fluentd_node2 = { + "metadata": { + "labels": {"logging-infra-fluentd": "true", "kubernetes.io/hostname": "hostname"}, + "name": "node2", + }, + "status": {"addresses": [{"type": "InternalIP", "address": "10.10.1.2"}]}, +} +fluentd_node3_unlabeled = { + "metadata": { + "labels": {"kubernetes.io/hostname": "hostname"}, + "name": "node3", + }, + "status": {"addresses": [{"type": "InternalIP", "address": "10.10.1.3"}]}, +} + + +def test_get_fluentd_pods(): + check = Fluentd() + check.exec_oc = lambda *_: json.dumps(dict(items=[fluentd_node1])) + check.get_pods_for_component = lambda *_: [fluentd_pod_node1] + assert not check.run() + + +@pytest.mark.parametrize('pods, nodes, expect_error', [ + ( + [], + [], + 'NoNodesDefined', + ), + ( + [], + [fluentd_node3_unlabeled], + 'NoNodesLabeled', + ), + ( + [], + [fluentd_node1, fluentd_node3_unlabeled], + 'NodesUnlabeled', + ), + ( + [], + [fluentd_node2], + 'MissingFluentdPod', + ), + ( + [fluentd_pod_node1, fluentd_pod_node1], + [fluentd_node1], + 'TooManyFluentdPods', + ), + ( + [fluentd_pod_node2_down], + [fluentd_node2], + 'FluentdNotRunning', + ), +]) +def test_get_fluentd_pods_errors(pods, nodes, expect_error): + check = Fluentd() + check.exec_oc = lambda *_: json.dumps(dict(items=nodes)) + + with pytest.raises(OpenShiftCheckException) as excinfo: + check.check_fluentd(pods) + if isinstance(excinfo.value, OpenShiftCheckExceptionList): + assert_error_in_list(expect_error, excinfo.value) + else: + assert expect_error == excinfo.value.name + + +def test_bad_oc_node_list(): + check = Fluentd() + check.exec_oc = lambda *_: "this isn't even json" + with pytest.raises(OpenShiftCheckException) as excinfo: + check.get_nodes_by_name() + assert 'BadOcNodeList' == excinfo.value.name diff --git a/roles/openshift_health_checker/test/kibana_test.py b/roles/openshift_health_checker/test/kibana_test.py new file mode 100644 index 000000000..04a5e89c4 --- /dev/null +++ b/roles/openshift_health_checker/test/kibana_test.py @@ -0,0 +1,244 @@ +import pytest +import json + +try: + import urllib2 + from urllib2 import HTTPError, URLError +except ImportError: + from urllib.error import HTTPError, URLError + import urllib.request as urllib2 + +from openshift_checks.logging.kibana import Kibana, OpenShiftCheckException + + +plain_kibana_pod = { + "metadata": { + "labels": {"component": "kibana", "deploymentconfig": "logging-kibana"}, + "name": "logging-kibana-1", + }, + "status": { + "containerStatuses": [{"ready": True}, {"ready": True}], + "conditions": [{"status": "True", "type": "Ready"}], + } +} +not_running_kibana_pod = { + "metadata": { + "labels": {"component": "kibana", "deploymentconfig": "logging-kibana"}, + "name": "logging-kibana-2", + }, + "status": { + "containerStatuses": [{"ready": True}, {"ready": False}], + "conditions": [{"status": "True", "type": "Ready"}], + } +} + + +def test_check_kibana(): + # should run without exception: + Kibana().check_kibana([plain_kibana_pod]) + + +@pytest.mark.parametrize('pods, expect_error', [ + ( + [], + "MissingComponentPods", + ), + ( + [not_running_kibana_pod], + "NoRunningPods", + ), + ( + [plain_kibana_pod, not_running_kibana_pod], + "PodNotRunning", + ), +]) +def test_check_kibana_error(pods, expect_error): + with pytest.raises(OpenShiftCheckException) as excinfo: + Kibana().check_kibana(pods) + assert expect_error == excinfo.value.name + + +@pytest.mark.parametrize('comment, route, expect_error', [ + ( + "No route returned", + None, + "no_route_exists", + ), + + ( + "broken route response", + {"status": {}}, + "get_route_failed", + ), + ( + "route with no ingress", + { + "metadata": { + "labels": {"component": "kibana", "deploymentconfig": "logging-kibana"}, + "name": "logging-kibana", + }, + "status": { + "ingress": [], + }, + "spec": { + "host": "hostname", + } + }, + "route_not_accepted", + ), + + ( + "route with no host", + { + "metadata": { + "labels": {"component": "kibana", "deploymentconfig": "logging-kibana"}, + "name": "logging-kibana", + }, + "status": { + "ingress": [{ + "status": True, + }], + }, + "spec": {}, + }, + "route_missing_host", + ), +]) +def test_get_kibana_url_error(comment, route, expect_error): + check = Kibana() + check.exec_oc = lambda *_: json.dumps(route) if route else "" + + with pytest.raises(OpenShiftCheckException) as excinfo: + check._get_kibana_url() + assert excinfo.value.name == expect_error + + +@pytest.mark.parametrize('comment, route, expect_url', [ + ( + "test route that looks fine", + { + "metadata": { + "labels": {"component": "kibana", "deploymentconfig": "logging-kibana"}, + "name": "logging-kibana", + }, + "status": { + "ingress": [{ + "status": True, + }], + }, + "spec": { + "host": "hostname", + }, + }, + "https://hostname/", + ), +]) +def test_get_kibana_url(comment, route, expect_url): + check = Kibana() + check.exec_oc = lambda *_: json.dumps(route) + assert expect_url == check._get_kibana_url() + + +@pytest.mark.parametrize('exec_result, expect', [ + ( + 'urlopen error [Errno 111] Connection refused', + 'FailedToConnectInternal', + ), + ( + 'urlopen error [Errno -2] Name or service not known', + 'FailedToResolveInternal', + ), + ( + 'Status code was not [302]: HTTP Error 500: Server error', + 'WrongReturnCodeInternal', + ), + ( + 'bork bork bork', + 'MiscRouteErrorInternal', + ), +]) +def test_verify_url_internal_failure(exec_result, expect): + check = Kibana(execute_module=lambda *_: dict(failed=True, msg=exec_result)) + check._get_kibana_url = lambda: 'url' + + with pytest.raises(OpenShiftCheckException) as excinfo: + check.check_kibana_route() + assert expect == excinfo.value.name + + +@pytest.mark.parametrize('lib_result, expect', [ + ( + HTTPError('url', 500, 'it broke', hdrs=None, fp=None), + 'MiscRouteError', + ), + ( + URLError('urlopen error [Errno 111] Connection refused'), + 'FailedToConnect', + ), + ( + URLError('urlopen error [Errno -2] Name or service not known'), + 'FailedToResolve', + ), + ( + 302, + 'WrongReturnCode', + ), + ( + 200, + None, + ), +]) +def test_verify_url_external_failure(lib_result, expect, monkeypatch): + + class _http_return: + + def __init__(self, code): + self.code = code + + def getcode(self): + return self.code + + def urlopen(url, context): + if type(lib_result) is int: + return _http_return(lib_result) + raise lib_result + monkeypatch.setattr(urllib2, 'urlopen', urlopen) + + check = Kibana() + check._get_kibana_url = lambda: 'url' + check._verify_url_internal = lambda url: None + + if not expect: + check.check_kibana_route() + return + + with pytest.raises(OpenShiftCheckException) as excinfo: + check.check_kibana_route() + assert expect == excinfo.value.name + + +def test_verify_url_external_skip(): + check = Kibana(lambda *_: {}, dict(openshift_check_efk_kibana_external="false")) + check._get_kibana_url = lambda: 'url' + check.check_kibana_route() + + +# this is kind of silly but it adds coverage for the run() method... +def test_run(): + pods = ["foo"] + ran = dict(check_kibana=False, check_route=False) + + def check_kibana(pod_list): + ran["check_kibana"] = True + assert pod_list == pods + + def check_kibana_route(): + ran["check_route"] = True + + check = Kibana() + check.get_pods_for_component = lambda *_: pods + check.check_kibana = check_kibana + check.check_kibana_route = check_kibana_route + + check.run() + assert ran["check_kibana"] and ran["check_route"] diff --git a/roles/openshift_health_checker/test/logging_check_test.py b/roles/openshift_health_checker/test/logging_check_test.py new file mode 100644 index 000000000..1a1c190f6 --- /dev/null +++ b/roles/openshift_health_checker/test/logging_check_test.py @@ -0,0 +1,168 @@ +import pytest +import json + +from openshift_checks.logging.logging import LoggingCheck, MissingComponentPods, CouldNotUseOc + +task_vars_config_base = dict(openshift=dict(common=dict(config_base='/etc/origin'))) + + +def canned_loggingcheck(exec_oc=None, execute_module=None): + """Create a LoggingCheck object with canned exec_oc method""" + check = LoggingCheck(execute_module) + if exec_oc: + check.exec_oc = exec_oc + return check + + +def assert_error(error, expect_error): + if expect_error: + assert error + assert expect_error in error + else: + assert not error + + +plain_es_pod = { + "metadata": { + "labels": {"component": "es", "deploymentconfig": "logging-es"}, + "name": "logging-es", + }, + "status": { + "conditions": [{"status": "True", "type": "Ready"}], + "containerStatuses": [{"ready": True}], + "podIP": "10.10.10.10", + }, + "_test_master_name_str": "name logging-es", +} + +plain_kibana_pod = { + "metadata": { + "labels": {"component": "kibana", "deploymentconfig": "logging-kibana"}, + "name": "logging-kibana-1", + }, + "status": { + "containerStatuses": [{"ready": True}, {"ready": True}], + "conditions": [{"status": "True", "type": "Ready"}], + } +} + +plain_kibana_pod_no_containerstatus = { + "metadata": { + "labels": {"component": "kibana", "deploymentconfig": "logging-kibana"}, + "name": "logging-kibana-1", + }, + "status": { + "conditions": [{"status": "True", "type": "Ready"}], + } +} + +fluentd_pod_node1 = { + "metadata": { + "labels": {"component": "fluentd", "deploymentconfig": "logging-fluentd"}, + "name": "logging-fluentd-1", + }, + "spec": {"host": "node1", "nodeName": "node1"}, + "status": { + "containerStatuses": [{"ready": True}], + "conditions": [{"status": "True", "type": "Ready"}], + } +} + +plain_curator_pod = { + "metadata": { + "labels": {"component": "curator", "deploymentconfig": "logging-curator"}, + "name": "logging-curator-1", + }, + "status": { + "containerStatuses": [{"ready": True}], + "conditions": [{"status": "True", "type": "Ready"}], + "podIP": "10.10.10.10", + } +} + + +@pytest.mark.parametrize('problem, expect', [ + ("[Errno 2] No such file or directory", "supposed to be a master"), + ("Permission denied", "Unexpected error using `oc`"), +]) +def test_oc_failure(problem, expect): + def execute_module(module_name, *_): + if module_name == "ocutil": + return dict(failed=True, result=problem) + return dict(changed=False) + + check = LoggingCheck(execute_module, task_vars_config_base) + + with pytest.raises(CouldNotUseOc) as excinfo: + check.exec_oc('get foo', []) + assert expect in str(excinfo) + + +groups_with_first_master = dict(masters=['this-host', 'other-host']) +groups_with_second_master = dict(masters=['other-host', 'this-host']) +groups_not_a_master = dict(masters=['other-host']) + + +@pytest.mark.parametrize('groups, logging_deployed, is_active', [ + (groups_with_first_master, True, True), + (groups_with_first_master, False, False), + (groups_not_a_master, True, False), + (groups_with_second_master, True, False), + (groups_not_a_master, True, False), +]) +def test_is_active(groups, logging_deployed, is_active): + task_vars = dict( + ansible_ssh_host='this-host', + groups=groups, + openshift_hosted_logging_deploy=logging_deployed, + ) + + assert LoggingCheck(None, task_vars).is_active() == is_active + + +@pytest.mark.parametrize('pod_output, expect_pods', [ + ( + json.dumps({'items': [plain_es_pod]}), + [plain_es_pod], + ), +]) +def test_get_pods_for_component(pod_output, expect_pods): + check = canned_loggingcheck(lambda *_: pod_output) + pods = check.get_pods_for_component("es") + assert pods == expect_pods + + +@pytest.mark.parametrize('exec_oc_output, expect_error', [ + ( + 'No resources found.', + MissingComponentPods, + ), + ( + '{"items": null}', + MissingComponentPods, + ), +]) +def test_get_pods_for_component_fail(exec_oc_output, expect_error): + check = canned_loggingcheck(lambda *_: exec_oc_output) + with pytest.raises(expect_error): + check.get_pods_for_component("es") + + +@pytest.mark.parametrize('name, pods, expected_pods', [ + ( + 'test single pod found, scheduled, but no containerStatuses field', + [plain_kibana_pod_no_containerstatus], + [plain_kibana_pod_no_containerstatus], + ), + ( + 'set of pods has at least one pod with containerStatuses (scheduled); should still fail', + [plain_kibana_pod_no_containerstatus, plain_kibana_pod], + [plain_kibana_pod_no_containerstatus], + ), + +], ids=lambda argvals: argvals[0]) +def test_get_not_running_pods_no_container_status(name, pods, expected_pods): + check = canned_loggingcheck(lambda *_: '') + result = check.not_running_pods(pods) + + assert result == expected_pods diff --git a/roles/openshift_health_checker/test/logging_index_time_test.py b/roles/openshift_health_checker/test/logging_index_time_test.py new file mode 100644 index 000000000..22566b295 --- /dev/null +++ b/roles/openshift_health_checker/test/logging_index_time_test.py @@ -0,0 +1,170 @@ +import json + +import pytest + +from openshift_checks.logging.logging_index_time import LoggingIndexTime, OpenShiftCheckException + + +SAMPLE_UUID = "unique-test-uuid" + + +def canned_loggingindextime(exec_oc=None): + """Create a check object with a canned exec_oc method""" + check = LoggingIndexTime() # fails if a module is actually invoked + if exec_oc: + check.exec_oc = exec_oc + return check + + +plain_running_elasticsearch_pod = { + "metadata": { + "labels": {"component": "es", "deploymentconfig": "logging-es-data-master"}, + "name": "logging-es-data-master-1", + }, + "status": { + "containerStatuses": [{"ready": True}, {"ready": True}], + "phase": "Running", + } +} +plain_running_kibana_pod = { + "metadata": { + "labels": {"component": "kibana", "deploymentconfig": "logging-kibana"}, + "name": "logging-kibana-1", + }, + "status": { + "containerStatuses": [{"ready": True}, {"ready": True}], + "phase": "Running", + } +} +not_running_kibana_pod = { + "metadata": { + "labels": {"component": "kibana", "deploymentconfig": "logging-kibana"}, + "name": "logging-kibana-2", + }, + "status": { + "containerStatuses": [{"ready": True}, {"ready": False}], + "conditions": [{"status": "True", "type": "Ready"}], + "phase": "pending", + } +} + + +@pytest.mark.parametrize('pods, expect_pods', [ + ( + [not_running_kibana_pod], + [], + ), + ( + [plain_running_kibana_pod], + [plain_running_kibana_pod], + ), + ( + [], + [], + ) +]) +def test_check_running_pods(pods, expect_pods): + check = canned_loggingindextime() + pods = check.running_pods(pods) + assert pods == expect_pods + + +def test_bad_config_param(): + with pytest.raises(OpenShiftCheckException) as error: + LoggingIndexTime(task_vars=dict(openshift_check_logging_index_timeout_seconds="foo")).run() + assert 'InvalidTimeout' == error.value.name + + +def test_no_running_pods(): + check = LoggingIndexTime() + check.get_pods_for_component = lambda *_: [not_running_kibana_pod] + with pytest.raises(OpenShiftCheckException) as error: + check.run() + assert 'kibanaNoRunningPods' == error.value.name + + +def test_with_running_pods(): + check = LoggingIndexTime() + check.get_pods_for_component = lambda *_: [plain_running_kibana_pod, plain_running_elasticsearch_pod] + check.curl_kibana_with_uuid = lambda *_: SAMPLE_UUID + check.wait_until_cmd_or_err = lambda *_: None + assert not check.run().get("failed") + + +@pytest.mark.parametrize('name, json_response, uuid, timeout', [ + ( + 'valid count in response', + { + "count": 1, + }, + SAMPLE_UUID, + 0.001, + ), +], ids=lambda argval: argval[0]) +def test_wait_until_cmd_or_err_succeeds(name, json_response, uuid, timeout): + check = canned_loggingindextime(lambda *_: json.dumps(json_response)) + check.wait_until_cmd_or_err(plain_running_elasticsearch_pod, uuid, timeout) + + +@pytest.mark.parametrize('name, json_response, timeout, expect_error', [ + ( + 'invalid json response', + { + "invalid_field": 1, + }, + 0.001, + 'esInvalidResponse', + ), + ( + 'empty response', + {}, + 0.001, + 'esInvalidResponse', + ), + ( + 'valid response but invalid match count', + { + "count": 0, + }, + 0.005, + 'NoMatchFound', + ) +], ids=lambda argval: argval[0]) +def test_wait_until_cmd_or_err(name, json_response, timeout, expect_error): + check = canned_loggingindextime(lambda *_: json.dumps(json_response)) + with pytest.raises(OpenShiftCheckException) as error: + check.wait_until_cmd_or_err(plain_running_elasticsearch_pod, SAMPLE_UUID, timeout) + + assert expect_error == error.value.name + + +def test_curl_kibana_with_uuid(): + check = canned_loggingindextime(lambda *_: json.dumps({"statusCode": 404})) + check.generate_uuid = lambda: SAMPLE_UUID + assert SAMPLE_UUID == check.curl_kibana_with_uuid(plain_running_kibana_pod) + + +@pytest.mark.parametrize('name, json_response, expect_error', [ + ( + 'invalid json response', + { + "invalid_field": "invalid", + }, + 'kibanaInvalidResponse', + ), + ( + 'wrong error code in response', + { + "statusCode": 500, + }, + 'kibanaInvalidReturnCode', + ), +], ids=lambda argval: argval[0]) +def test_failed_curl_kibana_with_uuid(name, json_response, expect_error): + check = canned_loggingindextime(lambda *_: json.dumps(json_response)) + check.generate_uuid = lambda: SAMPLE_UUID + + with pytest.raises(OpenShiftCheckException) as error: + check.curl_kibana_with_uuid(plain_running_kibana_pod) + + assert expect_error == error.value.name diff --git a/roles/openshift_health_checker/test/memory_availability_test.py b/roles/openshift_health_checker/test/memory_availability_test.py new file mode 100644 index 000000000..aee2f0416 --- /dev/null +++ b/roles/openshift_health_checker/test/memory_availability_test.py @@ -0,0 +1,127 @@ +import pytest + +from openshift_checks.memory_availability import MemoryAvailability + + +@pytest.mark.parametrize('group_names,is_active', [ + (['masters'], True), + (['nodes'], True), + (['etcd'], True), + (['masters', 'nodes'], True), + (['masters', 'etcd'], True), + ([], False), + (['lb'], False), + (['nfs'], False), +]) +def test_is_active(group_names, is_active): + task_vars = dict( + group_names=group_names, + ) + assert MemoryAvailability(None, task_vars).is_active() == is_active + + +@pytest.mark.parametrize('group_names,configured_min,ansible_memtotal_mb', [ + ( + ['masters'], + 0, + 17200, + ), + ( + ['nodes'], + 0, + 8200, + ), + ( + ['nodes'], + 1, # configure lower threshold + 2000, # too low for recommended but not for configured + ), + ( + ['nodes'], + 2, # configure threshold where adjustment pushes it over + 1900, + ), + ( + ['etcd'], + 0, + 8200, + ), + ( + ['masters', 'nodes'], + 0, + 17000, + ), +]) +def test_succeeds_with_recommended_memory(group_names, configured_min, ansible_memtotal_mb): + task_vars = dict( + group_names=group_names, + openshift_check_min_host_memory_gb=configured_min, + ansible_memtotal_mb=ansible_memtotal_mb, + ) + + result = MemoryAvailability(fake_execute_module, task_vars).run() + + assert not result.get('failed', False) + + +@pytest.mark.parametrize('group_names,configured_min,ansible_memtotal_mb,extra_words', [ + ( + ['masters'], + 0, + 0, + ['0.0 GiB'], + ), + ( + ['nodes'], + 0, + 100, + ['0.1 GiB'], + ), + ( + ['nodes'], + 24, # configure higher threshold + 20 * 1024, # enough to meet recommended but not configured + ['20.0 GiB'], + ), + ( + ['nodes'], + 24, # configure higher threshold + 22 * 1024, # not enough for adjustment to push over threshold + ['22.0 GiB'], + ), + ( + ['etcd'], + 0, + 6 * 1024, + ['6.0 GiB'], + ), + ( + ['etcd', 'masters'], + 0, + 9 * 1024, # enough memory for etcd, not enough for a master + ['9.0 GiB'], + ), + ( + ['nodes', 'masters'], + 0, + # enough memory for a node, not enough for a master + 11 * 1024, + ['11.0 GiB'], + ), +]) +def test_fails_with_insufficient_memory(group_names, configured_min, ansible_memtotal_mb, extra_words): + task_vars = dict( + group_names=group_names, + openshift_check_min_host_memory_gb=configured_min, + ansible_memtotal_mb=ansible_memtotal_mb, + ) + + result = MemoryAvailability(fake_execute_module, task_vars).run() + + assert result.get('failed', False) + for word in 'below recommended'.split() + extra_words: + assert word in result['msg'] + + +def fake_execute_module(*args): + raise AssertionError('this function should not be called') diff --git a/roles/openshift_health_checker/test/mixins_test.py b/roles/openshift_health_checker/test/mixins_test.py new file mode 100644 index 000000000..b1a41ca3c --- /dev/null +++ b/roles/openshift_health_checker/test/mixins_test.py @@ -0,0 +1,23 @@ +import pytest + +from openshift_checks import OpenShiftCheck, OpenShiftCheckException +from openshift_checks.mixins import NotContainerizedMixin + + +class NotContainerizedCheck(NotContainerizedMixin, OpenShiftCheck): + name = "not_containerized" + run = NotImplemented + + +@pytest.mark.parametrize('task_vars,expected', [ + (dict(openshift=dict(common=dict(is_containerized=False))), True), + (dict(openshift=dict(common=dict(is_containerized=True))), False), +]) +def test_is_active(task_vars, expected): + assert NotContainerizedCheck(None, task_vars).is_active() == expected + + +def test_is_active_missing_task_vars(): + with pytest.raises(OpenShiftCheckException) as excinfo: + NotContainerizedCheck().is_active() + assert 'is_containerized' in str(excinfo.value) diff --git a/roles/openshift_health_checker/test/openshift_check_test.py b/roles/openshift_health_checker/test/openshift_check_test.py new file mode 100644 index 000000000..789784c77 --- /dev/null +++ b/roles/openshift_health_checker/test/openshift_check_test.py @@ -0,0 +1,118 @@ +import pytest + +from openshift_checks import OpenShiftCheck, OpenShiftCheckException +from openshift_checks import load_checks + + +# Fixtures + + +@pytest.fixture() +def task_vars(): + return dict(foo=42, bar=dict(baz="openshift")) + + +@pytest.fixture(params=[ + ("notfound",), + ("multiple", "keys", "not", "in", "task_vars"), +]) +def missing_keys(request): + return request.param + + +# Tests + + +def test_OpenShiftCheck_init(): + class TestCheck(OpenShiftCheck): + name = "test_check" + run = NotImplemented + + # execute_module required at init if it will be used + with pytest.raises(RuntimeError) as excinfo: + TestCheck().execute_module("foo") + assert 'execute_module' in str(excinfo.value) + + execute_module = object() + + # initialize with positional argument + check = TestCheck(execute_module) + assert check._execute_module == execute_module + + # initialize with keyword argument + check = TestCheck(execute_module=execute_module) + assert check._execute_module == execute_module + + assert check.task_vars == {} + assert check.tmp is None + + +def test_subclasses(): + """OpenShiftCheck.subclasses should find all subclasses recursively.""" + class TestCheck1(OpenShiftCheck): + pass + + class TestCheck2(OpenShiftCheck): + pass + + class TestCheck1A(TestCheck1): + pass + + local_subclasses = set([TestCheck1, TestCheck1A, TestCheck2]) + known_subclasses = set(OpenShiftCheck.subclasses()) + + assert local_subclasses - known_subclasses == set(), "local_subclasses should be a subset of known_subclasses" + + +def test_load_checks(): + """Loading checks should load and return Python modules.""" + modules = load_checks() + assert modules + + +def dummy_check(task_vars): + class TestCheck(OpenShiftCheck): + name = "dummy" + run = NotImplemented + + return TestCheck(task_vars=task_vars) + + +@pytest.mark.parametrize("keys,expected", [ + (("foo",), 42), + (("bar", "baz"), "openshift"), + (("bar.baz",), "openshift"), +]) +def test_get_var_ok(task_vars, keys, expected): + assert dummy_check(task_vars).get_var(*keys) == expected + + +def test_get_var_error(task_vars, missing_keys): + with pytest.raises(OpenShiftCheckException): + dummy_check(task_vars).get_var(*missing_keys) + + +def test_get_var_default(task_vars, missing_keys): + default = object() + assert dummy_check(task_vars).get_var(*missing_keys, default=default) == default + + +@pytest.mark.parametrize("keys, convert, expected", [ + (("foo",), str, "42"), + (("foo",), float, 42.0), + (("bar", "baz"), bool, False), +]) +def test_get_var_convert(task_vars, keys, convert, expected): + assert dummy_check(task_vars).get_var(*keys, convert=convert) == expected + + +@pytest.mark.parametrize("keys, convert", [ + (("bar", "baz"), int), + (("bar.baz"), float), + (("foo"), "bogus"), + (("foo"), lambda a, b: 1), + (("foo"), lambda a: 1 / 0), +]) +def test_get_var_convert_error(task_vars, keys, convert): + with pytest.raises(OpenShiftCheckException): + dummy_check(task_vars).get_var(*keys, convert=convert) diff --git a/roles/openshift_health_checker/test/ovs_version_test.py b/roles/openshift_health_checker/test/ovs_version_test.py new file mode 100644 index 000000000..e1bf29d2a --- /dev/null +++ b/roles/openshift_health_checker/test/ovs_version_test.py @@ -0,0 +1,86 @@ +import pytest + +from openshift_checks.ovs_version import OvsVersion, OpenShiftCheckException + + +def test_openshift_version_not_supported(): + def execute_module(*_): + return {} + + openshift_release = '111.7.0' + + task_vars = dict( + openshift=dict(common=dict(service_type='origin')), + openshift_release=openshift_release, + openshift_image_tag='v' + openshift_release, + openshift_deployment_type='origin', + ) + + with pytest.raises(OpenShiftCheckException) as excinfo: + OvsVersion(execute_module, task_vars).run() + + assert "no recommended version of Open vSwitch" in str(excinfo.value) + + +def test_invalid_openshift_release_format(): + def execute_module(*_): + return {} + + task_vars = dict( + openshift=dict(common=dict(service_type='origin')), + openshift_image_tag='v0', + openshift_deployment_type='origin', + ) + + with pytest.raises(OpenShiftCheckException) as excinfo: + OvsVersion(execute_module, task_vars).run() + assert "invalid version" in str(excinfo.value) + + +@pytest.mark.parametrize('openshift_release,expected_ovs_version', [ + ("3.5", ["2.6", "2.7"]), + ("3.6", ["2.6", "2.7"]), + ("3.4", "2.4"), + ("3.3", "2.4"), + ("1.0", "2.4"), +]) +def test_ovs_package_version(openshift_release, expected_ovs_version): + task_vars = dict( + openshift=dict(common=dict(service_type='origin')), + openshift_release=openshift_release, + openshift_image_tag='v' + openshift_release, + ) + return_value = object() + + def execute_module(module_name=None, module_args=None, *_): + assert module_name == 'rpm_version' + assert "package_list" in module_args + + for pkg in module_args["package_list"]: + if pkg["name"] == "openvswitch": + assert pkg["version"] == expected_ovs_version + + return return_value + + result = OvsVersion(execute_module, task_vars).run() + assert result is return_value + + +@pytest.mark.parametrize('group_names,is_containerized,is_active', [ + (['masters'], False, True), + # ensure check is skipped on containerized installs + (['masters'], True, False), + (['nodes'], False, True), + (['masters', 'nodes'], False, True), + (['masters', 'etcd'], False, True), + ([], False, False), + (['etcd'], False, False), + (['lb'], False, False), + (['nfs'], False, False), +]) +def test_ovs_version_skip_when_not_master_nor_node(group_names, is_containerized, is_active): + task_vars = dict( + group_names=group_names, + openshift=dict(common=dict(is_containerized=is_containerized)), + ) + assert OvsVersion(None, task_vars).is_active() == is_active diff --git a/roles/openshift_health_checker/test/package_availability_test.py b/roles/openshift_health_checker/test/package_availability_test.py new file mode 100644 index 000000000..1fe648b75 --- /dev/null +++ b/roles/openshift_health_checker/test/package_availability_test.py @@ -0,0 +1,62 @@ +import pytest + +from openshift_checks.package_availability import PackageAvailability + + +@pytest.mark.parametrize('pkg_mgr,is_containerized,is_active', [ + ('yum', False, True), + ('yum', True, False), + ('dnf', True, False), + ('dnf', False, False), +]) +def test_is_active(pkg_mgr, is_containerized, is_active): + task_vars = dict( + ansible_pkg_mgr=pkg_mgr, + openshift=dict(common=dict(is_containerized=is_containerized)), + ) + assert PackageAvailability(None, task_vars).is_active() == is_active + + +@pytest.mark.parametrize('task_vars,must_have_packages,must_not_have_packages', [ + ( + dict(openshift=dict(common=dict(service_type='openshift'))), + set(), + set(['openshift-master', 'openshift-node']), + ), + ( + dict( + openshift=dict(common=dict(service_type='origin')), + group_names=['masters'], + ), + set(['origin-master']), + set(['origin-node']), + ), + ( + dict( + openshift=dict(common=dict(service_type='atomic-openshift')), + group_names=['nodes'], + ), + set(['atomic-openshift-node']), + set(['atomic-openshift-master']), + ), + ( + dict( + openshift=dict(common=dict(service_type='atomic-openshift')), + group_names=['masters', 'nodes'], + ), + set(['atomic-openshift-master', 'atomic-openshift-node']), + set(), + ), +]) +def test_package_availability(task_vars, must_have_packages, must_not_have_packages): + return_value = object() + + def execute_module(module_name=None, module_args=None, *_): + assert module_name == 'check_yum_update' + assert 'packages' in module_args + assert set(module_args['packages']).issuperset(must_have_packages) + assert not set(module_args['packages']).intersection(must_not_have_packages) + return return_value + + result = PackageAvailability(execute_module, task_vars).run() + assert result is return_value diff --git a/roles/openshift_health_checker/test/package_update_test.py b/roles/openshift_health_checker/test/package_update_test.py new file mode 100644 index 000000000..06489b0d7 --- /dev/null +++ b/roles/openshift_health_checker/test/package_update_test.py @@ -0,0 +1,15 @@ +from openshift_checks.package_update import PackageUpdate + + +def test_package_update(): + return_value = object() + + def execute_module(module_name=None, module_args=None, *_): + assert module_name == 'check_yum_update' + assert 'packages' in module_args + # empty list of packages means "generic check if 'yum update' will work" + assert module_args['packages'] == [] + return return_value + + result = PackageUpdate(execute_module).run() + assert result is return_value diff --git a/roles/openshift_health_checker/test/package_version_test.py b/roles/openshift_health_checker/test/package_version_test.py new file mode 100644 index 000000000..6054d3f3e --- /dev/null +++ b/roles/openshift_health_checker/test/package_version_test.py @@ -0,0 +1,114 @@ +import pytest + +from openshift_checks.package_version import PackageVersion, OpenShiftCheckException + + +def task_vars_for(openshift_release, deployment_type): + return dict( + openshift=dict(common=dict(service_type=deployment_type)), + openshift_release=openshift_release, + openshift_image_tag='v' + openshift_release, + openshift_deployment_type=deployment_type, + ) + + +def test_openshift_version_not_supported(): + check = PackageVersion(None, task_vars_for("1.2.3", 'origin')) + check.get_openshift_version_tuple = lambda: (3, 4, 1) # won't be in the dict + + with pytest.raises(OpenShiftCheckException) as excinfo: + check.get_required_ovs_version() + assert "no recommended version of Open vSwitch" in str(excinfo.value) + + with pytest.raises(OpenShiftCheckException) as excinfo: + check.get_required_docker_version() + assert "no recommended version of Docker" in str(excinfo.value) + + +def test_invalid_openshift_release_format(): + task_vars = dict( + openshift=dict(common=dict(service_type='origin')), + openshift_image_tag='v0', + openshift_deployment_type='origin', + ) + + check = PackageVersion(lambda *_: {}, task_vars) + with pytest.raises(OpenShiftCheckException) as excinfo: + check.run() + assert "invalid version" in str(excinfo.value) + + +@pytest.mark.parametrize('openshift_release', [ + "111.7.0", + "3.7", + "3.6", + "3.5.1.2.3", + "3.5", + "3.4", + "3.3", + "2.1.0", +]) +def test_package_version(openshift_release): + + return_value = object() + + def execute_module(module_name=None, module_args=None, tmp=None, task_vars=None, *_): + assert module_name == 'aos_version' + assert "package_list" in module_args + + for pkg in module_args["package_list"]: + if "-master" in pkg["name"] or "-node" in pkg["name"]: + assert pkg["version"] == task_vars["openshift_release"] + + return return_value + + check = PackageVersion(execute_module, task_vars_for(openshift_release, 'origin')) + result = check.run() + assert result is return_value + + +@pytest.mark.parametrize('deployment_type,openshift_release,expected_docker_version', [ + ("origin", "3.5", "1.12"), + ("origin", "1.3", "1.10"), + ("origin", "1.1", "1.8"), + ("openshift-enterprise", "3.4", "1.12"), + ("openshift-enterprise", "3.2", "1.10"), + ("openshift-enterprise", "3.1", "1.8"), +]) +def test_docker_package_version(deployment_type, openshift_release, expected_docker_version): + + return_value = object() + + def execute_module(module_name=None, module_args=None, *_): + assert module_name == 'aos_version' + assert "package_list" in module_args + + for pkg in module_args["package_list"]: + if pkg["name"] == "docker": + assert pkg["version"] == expected_docker_version + + return return_value + + check = PackageVersion(execute_module, task_vars_for(openshift_release, deployment_type)) + result = check.run() + assert result is return_value + + +@pytest.mark.parametrize('group_names,is_containerized,is_active', [ + (['masters'], False, True), + # ensure check is skipped on containerized installs + (['masters'], True, False), + (['nodes'], False, True), + (['masters', 'nodes'], False, True), + (['masters', 'etcd'], False, True), + ([], False, False), + (['etcd'], False, False), + (['lb'], False, False), + (['nfs'], False, False), +]) +def test_package_version_skip_when_not_master_nor_node(group_names, is_containerized, is_active): + task_vars = dict( + group_names=group_names, + openshift=dict(common=dict(is_containerized=is_containerized)), + ) + assert PackageVersion(None, task_vars).is_active() == is_active diff --git a/roles/openshift_health_checker/test/rpm_version_test.py b/roles/openshift_health_checker/test/rpm_version_test.py new file mode 100644 index 000000000..2c1bcf876 --- /dev/null +++ b/roles/openshift_health_checker/test/rpm_version_test.py @@ -0,0 +1,82 @@ +import pytest +import rpm_version + +expected_pkgs = { + "spam": { + "name": "spam", + "version": "3.2.1", + }, + "eggs": { + "name": "eggs", + "version": "3.2.1", + }, +} + + +@pytest.mark.parametrize('pkgs, expect_not_found', [ + ( + {}, + ["spam", "eggs"], # none found + ), + ( + {"spam": ["3.2.1", "4.5.1"]}, + ["eggs"], # completely missing + ), + ( + { + "spam": ["3.2.1", "4.5.1"], + "eggs": ["3.2.1"], + }, + [], # all found + ), +]) +def test_check_pkg_found(pkgs, expect_not_found): + if expect_not_found: + with pytest.raises(rpm_version.RpmVersionException) as e: + rpm_version._check_pkg_versions(pkgs, expected_pkgs) + + assert "not found to be installed" in str(e.value) + assert set(expect_not_found) == set(e.value.problem_pkgs) + else: + rpm_version._check_pkg_versions(pkgs, expected_pkgs) + + +@pytest.mark.parametrize('pkgs, expect_not_found', [ + ( + { + 'spam': ['3.2.1'], + 'eggs': ['3.3.2'], + }, + { + "eggs": { + "required_versions": ["3.2"], + "found_versions": ["3.3"], + } + }, # not the right version + ), + ( + { + 'spam': ['3.1.2', "3.3.2"], + 'eggs': ['3.3.2', "1.2.3"], + }, + { + "eggs": { + "required_versions": ["3.2"], + "found_versions": ["3.3", "1.2"], + }, + "spam": { + "required_versions": ["3.2"], + "found_versions": ["3.1", "3.3"], + } + }, # not the right version + ), +]) +def test_check_pkg_version_found(pkgs, expect_not_found): + if expect_not_found: + with pytest.raises(rpm_version.RpmVersionException) as e: + rpm_version._check_pkg_versions(pkgs, expected_pkgs) + + assert "found to be installed with an incorrect version" in str(e.value) + assert expect_not_found == e.value.problem_pkgs + else: + rpm_version._check_pkg_versions(pkgs, expected_pkgs) diff --git a/roles/openshift_health_checker/test/search_journalctl_test.py b/roles/openshift_health_checker/test/search_journalctl_test.py new file mode 100644 index 000000000..724928aa1 --- /dev/null +++ b/roles/openshift_health_checker/test/search_journalctl_test.py @@ -0,0 +1,157 @@ +import pytest +import search_journalctl + + +def canned_search_journalctl(get_log_output=None): + """Create a search_journalctl object with canned get_log_output method""" + module = search_journalctl + if get_log_output: + module.get_log_output = get_log_output + return module + + +DEFAULT_TIMESTAMP = 1496341364 + + +def get_timestamp(modifier=0): + return DEFAULT_TIMESTAMP + modifier + + +def get_timestamp_microseconds(modifier=0): + return get_timestamp(modifier) * 1000000 + + +def create_test_log_object(stamp, msg): + return '{{"__REALTIME_TIMESTAMP": "{}", "MESSAGE": "{}"}}'.format(stamp, msg) + + +@pytest.mark.parametrize('name,matchers,log_input,expected_matches,expected_errors', [ + ( + 'test with valid params', + [ + { + "start_regexp": r"Sample Logs Beginning", + "regexp": r"test log message", + "unit": "test", + }, + ], + [ + create_test_log_object(get_timestamp_microseconds(), "test log message"), + create_test_log_object(get_timestamp_microseconds(), "Sample Logs Beginning"), + ], + ["test log message"], + [], + ), + ( + 'test with invalid json in log input', + [ + { + "start_regexp": r"Sample Logs Beginning", + "regexp": r"test log message", + "unit": "test-unit", + }, + ], + [ + '{__REALTIME_TIMESTAMP: ' + str(get_timestamp_microseconds()) + ', "MESSAGE": "test log message"}', + ], + [], + [ + ["invalid json", "test-unit", "test log message"], + ], + ), + ( + 'test with invalid regexp', + [ + { + "start_regexp": r"Sample Logs Beginning", + "regexp": r"test [ log message", + "unit": "test", + }, + ], + [ + create_test_log_object(get_timestamp_microseconds(), "test log message"), + create_test_log_object(get_timestamp_microseconds(), "sample log message"), + create_test_log_object(get_timestamp_microseconds(), "fake log message"), + create_test_log_object(get_timestamp_microseconds(), "dummy log message"), + create_test_log_object(get_timestamp_microseconds(), "Sample Logs Beginning"), + ], + [], + [ + ["invalid regular expression"], + ], + ), +], ids=lambda argval: argval[0]) +def test_get_log_matches(name, matchers, log_input, expected_matches, expected_errors): + def get_log_output(matcher): + return log_input + + module = canned_search_journalctl(get_log_output) + matched_regexp, errors = module.get_log_matches(matchers, 500, 60 * 60) + + assert set(matched_regexp) == set(expected_matches) + assert len(expected_errors) == len(errors) + + for idx, partial_err_set in enumerate(expected_errors): + for partial_err_msg in partial_err_set: + assert partial_err_msg in errors[idx] + + +@pytest.mark.parametrize('name,matcher,log_count_lim,stamp_lim_seconds,log_input,expected_match', [ + ( + 'test with matching log message, but out of bounds of log_count_lim', + { + "start_regexp": r"Sample Logs Beginning", + "regexp": r"dummy log message", + "unit": "test", + }, + 3, + get_timestamp(-100 * 60 * 60), + [ + create_test_log_object(get_timestamp_microseconds(), "test log message"), + create_test_log_object(get_timestamp_microseconds(), "sample log message"), + create_test_log_object(get_timestamp_microseconds(), "fake log message"), + create_test_log_object(get_timestamp_microseconds(), "dummy log message"), + create_test_log_object(get_timestamp_microseconds(), "Sample Logs Beginning"), + ], + None, + ), + ( + 'test with matching log message, but with timestamp too old', + { + "start_regexp": r"Sample Logs Beginning", + "regexp": r"dummy log message", + "unit": "test", + }, + 100, + get_timestamp(-10), + [ + create_test_log_object(get_timestamp_microseconds(), "test log message"), + create_test_log_object(get_timestamp_microseconds(), "sample log message"), + create_test_log_object(get_timestamp_microseconds(), "fake log message"), + create_test_log_object(get_timestamp_microseconds(-1000), "dummy log message"), + create_test_log_object(get_timestamp_microseconds(-1000), "Sample Logs Beginning"), + ], + None, + ), + ( + 'test with matching log message, and timestamp within time limit', + { + "start_regexp": r"Sample Logs Beginning", + "regexp": r"dummy log message", + "unit": "test", + }, + 100, + get_timestamp(-1010), + [ + create_test_log_object(get_timestamp_microseconds(), "test log message"), + create_test_log_object(get_timestamp_microseconds(), "sample log message"), + create_test_log_object(get_timestamp_microseconds(), "fake log message"), + create_test_log_object(get_timestamp_microseconds(-1000), "dummy log message"), + create_test_log_object(get_timestamp_microseconds(-1000), "Sample Logs Beginning"), + ], + create_test_log_object(get_timestamp_microseconds(-1000), "dummy log message"), + ), +], ids=lambda argval: argval[0]) +def test_find_matches_skips_logs(name, matcher, log_count_lim, stamp_lim_seconds, log_input, expected_match): + match = search_journalctl.find_matches(log_input, matcher, log_count_lim, stamp_lim_seconds) + assert match == expected_match diff --git a/roles/openshift_health_checker/test/zz_failure_summary_test.py b/roles/openshift_health_checker/test/zz_failure_summary_test.py new file mode 100644 index 000000000..0fc258133 --- /dev/null +++ b/roles/openshift_health_checker/test/zz_failure_summary_test.py @@ -0,0 +1,70 @@ +from zz_failure_summary import deduplicate_failures + +import pytest + + +@pytest.mark.parametrize('failures,deduplicated', [ + ( + [ + { + 'host': 'master1', + 'msg': 'One or more checks failed', + }, + ], + [ + { + 'host': ('master1',), + 'msg': 'One or more checks failed', + }, + ], + ), + ( + [ + { + 'host': 'master1', + 'msg': 'One or more checks failed', + }, + { + 'host': 'node1', + 'msg': 'One or more checks failed', + }, + ], + [ + { + 'host': ('master1', 'node1'), + 'msg': 'One or more checks failed', + }, + ], + ), + ( + [ + { + 'host': 'node1', + 'msg': 'One or more checks failed', + 'checks': (('test_check', 'error message'),), + }, + { + 'host': 'master2', + 'msg': 'Some error happened', + }, + { + 'host': 'master1', + 'msg': 'One or more checks failed', + 'checks': (('test_check', 'error message'),), + }, + ], + [ + { + 'host': ('master1', 'node1'), + 'msg': 'One or more checks failed', + 'checks': (('test_check', 'error message'),), + }, + { + 'host': ('master2',), + 'msg': 'Some error happened', + }, + ], + ), +]) +def test_deduplicate_failures(failures, deduplicated): + assert deduplicate_failures(failures) == deduplicated |