Fix client tls (#239)

src/alertmanager_client.py

@@ -19,7 +19,7 @@
 logger = logging.getLogger(__name__)
 
 
-class AlertmanagerBadResponse(RuntimeError):
+class AlertmanagerBadResponse(ConnectionError):
     """A catch-all exception type to indicate 'no reply', regardless the reason."""
 
 

src/charm.py

@@ -19,7 +19,6 @@
     WorkloadManager,
     WorkloadManagerError,
 )
-from alertmanager_client import Alertmanager, AlertmanagerBadResponse
 from charms.alertmanager_k8s.v0.alertmanager_remote_configuration import (
     RemoteConfigurationRequirer,
 )
@@ -67,12 +66,7 @@
     ),
 )
 class AlertmanagerCharm(CharmBase):
-    """A Juju charm for alertmanager.
-
-    Attributes:
-        api: an API client instance for communicating with the alertmanager workload
-                server
-    """
+    """A Juju charm for alertmanager."""
 
     # Container name must match metadata.yaml
     # Layer name is used for the layer label argument in container.add_layer
@@ -124,8 +118,6 @@ def __init__(self, *args):
             external_url=self._internal_url,  # TODO See 'TODO' below, about external_url
         )
 
-        self.api = Alertmanager(endpoint_url=self._external_url)
-
         self.grafana_dashboard_provider = GrafanaDashboardProvider(charm=self)
         self.grafana_source_provider = GrafanaSourceProvider(
             charm=self,
@@ -498,7 +490,7 @@ def _on_update_status(self, _):
         Logs list of peers, uptime and version info.
         """
         try:
-            status = self.api.status()
+            status = self.alertmanager_workload.api.status()
             logger.info(
                 "alertmanager %s is up and running (uptime: %s); "
                 "cluster mode: %s, with %d peers",
@@ -507,7 +499,7 @@ def _on_update_status(self, _):
                 status["cluster"]["status"],
                 len(status["cluster"]["peers"]),
             )
-        except AlertmanagerBadResponse as e:
+        except ConnectionError as e:
             logger.error("Failed to obtain status: %s", str(e))
 
         # Calling the common hook to make sure a single unit set its IP in case all events fired

tests/integration/test_persistence.py

@@ -26,7 +26,10 @@ async def test_silences_persist_across_upgrades(ops_test: OpsTest, charm_under_t
     await ops_test.model.deploy(
         "ch:alertmanager-k8s", application_name=app_name, channel="edge", trust=True
     )
-    await ops_test.model.wait_for_idle(apps=[app_name], status="active", timeout=1000)
+    await ops_test.model.wait_for_idle(
+        apps=[app_name], status="active", timeout=1000, raise_on_error=False
+    )
+    await ops_test.model.wait_for_idle(apps=[app_name], status="active", timeout=30)
 
     # set a silencer for an alert and check it is set
     unit_address = await get_unit_address(ops_test, app_name, 0)

tests/integration/test_templates.py

@@ -5,13 +5,12 @@
 import json
 import logging
 import time
-from datetime import datetime, timedelta, timezone
 from pathlib import Path
 
 import pytest
+import sh
 import yaml
-from alertmanager_client import Alertmanager
-from helpers import get_unit_address, is_alertmanager_up
+from helpers import is_alertmanager_up
 from pytest_operator.plugin import OpsTest
 from werkzeug.wrappers import Request, Response
 
@@ -73,29 +72,6 @@ async def test_configure_alertmanager_with_templates(ops_test: OpsTest, httpserv
 
 @pytest.mark.abort_on_fail
 async def test_receiver_gets_alert(ops_test: OpsTest, httpserver):
-    # create an alert
-    start_time = datetime.now(timezone.utc)
-    end_time = start_time + timedelta(minutes=5)
-    alert_name = "fake-alert"
-    model_uuid = "1234"
-    alerts = [
-        {
-            "startsAt": start_time.isoformat("T"),
-            "endsAt": end_time.isoformat("T"),
-            "status": "firing",
-            "annotations": {
-                "summary": "A fake alert",
-            },
-            "labels": {
-                "juju_model_uuid": model_uuid,
-                "juju_application": app_name,
-                "juju_model": ops_test.model_name,
-                "alertname": alert_name,
-            },
-            "generatorURL": f"http://localhost/{alert_name}",
-        }
-    ]
-
     request_from_alertmanager = None
 
     def request_handler(request: Request):
@@ -127,9 +103,28 @@ def request_handler(request: Request):
     with httpserver.wait(timeout=120) as waiting:
         # expect an alert to be forwarded to the receiver
         httpserver.expect_oneshot_request("/", method="POST").respond_with_handler(request_handler)
-        unit_address = await get_unit_address(ops_test, app_name, 0)
-        amanager = Alertmanager(f"http://{unit_address}:9093")
-        amanager.set_alerts(alerts)
+
+        # Use amtool to fire a stand-in alert
+        sh.juju(
+            [
+                "ssh",
+                "-m",
+                ops_test.model_name,
+                "--container",
+                "alertmanager",
+                f"{app_name}/0",
+                "amtool",
+                "alert",
+                "add",
+                "foo",
+                "node=bar",
+                "status=firing",
+                "juju_model_uuid=1234",
+                f"juju_application={app_name}",
+                "juju_model=model_name",
+                "--annotation=summary=summary",
+            ]
+        )
 
     # check receiver got an alert
     assert waiting.result

tests/integration/test_upgrade_charm.py

@@ -71,7 +71,10 @@ async def test_upgrade_local_with_local_with_relations(ops_test: OpsTest, charm_
     # Refresh from path
     await ops_test.model.applications[app_name].refresh(path=charm_under_test, resources=resources)
     await ops_test.model.wait_for_idle(
-        apps=[app_name, "prom", "karma"], status="active", timeout=2500
+        apps=[app_name, "prom", "karma"],
+        status="active",
+        timeout=2500,
+        raise_on_error=False,
     )
     assert await is_alertmanager_up(ops_test, app_name)
 

tests/unit/test_charm.py

@@ -8,8 +8,8 @@
 import ops
 import yaml
 from alertmanager import WorkloadManager
-from charm import Alertmanager, AlertmanagerCharm
-from helpers import k8s_resource_multipatch, tautology
+from charm import AlertmanagerCharm
+from helpers import k8s_resource_multipatch
 from ops import pebble
 from ops.model import ActiveStatus, BlockedStatus
 from ops.testing import Harness
@@ -20,7 +20,6 @@
 class TestWithInitialHooks(unittest.TestCase):
     container_name: str = "alertmanager"
 
-    @patch.object(Alertmanager, "reload", tautology)
     @patch.object(WorkloadManager, "check_config", lambda *a, **kw: ("ok", ""))
     @patch("socket.getfqdn", new=lambda *args: "fqdn")
     @k8s_resource_multipatch
@@ -151,7 +150,6 @@ def test_templates_section_added_if_user_provided_templates(self, *unused):
 class TestWithoutInitialHooks(unittest.TestCase):
     container_name: str = "alertmanager"
 
-    @patch.object(Alertmanager, "reload", tautology)
     @patch.object(WorkloadManager, "check_config", lambda *a, **kw: ("ok", ""))
     @k8s_resource_multipatch
     @patch("lightkube.core.client.GenericSyncClient")

tests/unit/test_external_url.py

@@ -9,8 +9,8 @@
 import ops
 import yaml
 from alertmanager import WorkloadManager
-from charm import Alertmanager, AlertmanagerCharm
-from helpers import cli_arg, k8s_resource_multipatch, tautology
+from charm import AlertmanagerCharm
+from helpers import cli_arg, k8s_resource_multipatch
 from ops.testing import Harness
 
 logger = logging.getLogger(__name__)
@@ -21,7 +21,6 @@
 
 
 class TestExternalUrl(unittest.TestCase):
-    @patch.object(Alertmanager, "reload", tautology)
     @patch.object(WorkloadManager, "check_config", lambda *a, **kw: ("ok", ""))
     @patch("socket.getfqdn", new=lambda *args: "fqdn")
     @k8s_resource_multipatch

tests/unit/test_push_config_to_workload_on_startup.py

@@ -11,8 +11,8 @@
 import validators
 import yaml
 from alertmanager import WorkloadManager
-from charm import Alertmanager, AlertmanagerCharm
-from helpers import k8s_resource_multipatch, tautology
+from charm import AlertmanagerCharm
+from helpers import k8s_resource_multipatch
 from hypothesis import given
 from ops.model import ActiveStatus, BlockedStatus
 from ops.testing import Harness
@@ -29,7 +29,6 @@ class TestPushConfigToWorkloadOnStartup(unittest.TestCase):
     Background: Charm starts up with initial hooks.
     """
 
-    @patch.object(Alertmanager, "reload", tautology)
     @patch.object(WorkloadManager, "check_config", lambda *a, **kw: ("0.0.0", ""))
     @k8s_resource_multipatch
     @patch("lightkube.core.client.GenericSyncClient")
@@ -118,7 +117,6 @@ def setUp(self):
         self.harness = Harness(AlertmanagerCharm)
         self.addCleanup(self.harness.cleanup)
 
-    @patch.object(Alertmanager, "reload", tautology)
     @k8s_resource_multipatch
     @patch("lightkube.core.client.GenericSyncClient")
     @patch.object(WorkloadManager, "_alertmanager_version", property(lambda *_: "0.0.0"))
@@ -132,7 +130,6 @@ def test_charm_blocks_on_invalid_config_on_startup(self, *_):
         # THEN the charm goes into blocked status
         self.assertIsInstance(self.harness.charm.unit.status, BlockedStatus)
 
-    @patch.object(Alertmanager, "reload", tautology)
     @k8s_resource_multipatch
     @patch("lightkube.core.client.GenericSyncClient")
     @patch.object(WorkloadManager, "_alertmanager_version", property(lambda *_: "0.0.0"))

tox.ini

@@ -123,6 +123,7 @@ deps =
     pytest
     pytest-operator
     pytest-httpserver
+    sh
 commands =
     pytest -v --tb native --log-cli-level=INFO -s {posargs} {toxinidir}/tests/integration