feat: pd service level ops

Signed-off-by: 35C4n0r <[email protected]>

docs/providers/documentation/pagerduty-provider.mdx

@@ -30,6 +30,7 @@ To connect Keep to PagerDuty:
 
 - **Routing Key**: Use for event posting via the PagerDuty Events API.
 - **API Key**: Use for incident creation and management through the PagerDuty Incidents API.
+- **Service Id** (Optional): If provided, keep operates within the service's scope.
 - **OAuth2**: Token management handled automatically by Keep.
 
 <Frame>

keep/api/models/alert.py

@@ -517,6 +517,7 @@ def from_db_incident(cls, db_incident: "Incident"):
             assignee=db_incident.assignee,
             services=db_incident.affected_services or [],
             rule_fingerprint=db_incident.rule_fingerprint,
+            fingerprint=db_incident.fingerprint,
             same_incident_in_the_past_id=db_incident.same_incident_in_the_past_id,
             merged_into_incident_id=db_incident.merged_into_incident_id,
             merged_by=db_incident.merged_by,
@@ -550,6 +551,7 @@ def to_db_incident(self) -> "Incident":
             is_predicted=self.is_predicted,
             is_confirmed=self.is_confirmed,
             rule_fingerprint=self.rule_fingerprint,
+            fingerprint=self.fingerprint,
             same_incident_in_the_past_id=self.same_incident_in_the_past_id,
             merged_into_incident_id=self.merged_into_incident_id,
             merged_by=self.merged_by,

keep/providers/pagerduty_provider/pagerduty_provider.py

@@ -29,6 +29,7 @@
 from keep.providers.models.provider_config import ProviderConfig, ProviderScope
 from keep.providers.providers_factory import ProvidersFactory
 
+
 # Todo: think about splitting in to PagerdutyIncidentsProvider and PagerdutyAlertsProvider
 # Read this: https://community.pagerduty.com/forum/t/create-incident-using-python/3596/3
 
@@ -59,6 +60,14 @@ class PagerdutyProviderAuthConfig:
         },
         default="",
     )
+    service_id: str | None = dataclasses.field(
+        metadata={
+            "required": False,
+            "description": "Service Id (if provided, keep will only operate on this service)",
+            "sensitive": False,
+        },
+        default=None,
+    )
 
 
 class PagerdutyProvider(BaseTopologyProvider, BaseIncidentProvider):
@@ -134,7 +143,7 @@ class PagerdutyProvider(BaseTopologyProvider, BaseIncidentProvider):
     FINGERPRINT_FIELDS = ["alert_key"]
 
     def __init__(
-        self, context_manager: ContextManager, provider_id: str, config: ProviderConfig
+            self, context_manager: ContextManager, provider_id: str, config: ProviderConfig
     ):
         super().__init__(context_manager, provider_id, config)
 
@@ -155,7 +164,7 @@ def __init__(
             self.logger.info("Refreshing access token")
             self.__refresh_token()
         elif (
-            self.authentication_config.api_key or self.authentication_config.routing_key
+                self.authentication_config.api_key or self.authentication_config.routing_key
         ):
             # No need to do anything
             return
@@ -197,9 +206,9 @@ def validate_config(self):
             **self.config.authentication
         )
         if (
-            not self.authentication_config.routing_key
-            and not self.authentication_config.api_key
-            and not self.authentication_config.oauth_data
+                not self.authentication_config.routing_key
+                and not self.authentication_config.api_key
+                and not self.authentication_config.oauth_data
         ):
             raise ProviderConfigException(
                 "PagerdutyProvider requires either routing_key or api_key or OAuth configuration",
@@ -325,13 +334,13 @@ def validate_scopes(self):
         return scopes
 
     def _build_alert(
-        self,
-        title: str,
-        alert_body: str,
-        dedup: str | None = None,
-        severity: typing.Literal["critical", "error", "warning", "info"] | None = None,
-        event_type: typing.Literal["trigger", "acknowledge", "resolve"] | None = None,
-        source: str = "custom_event",
+            self,
+            title: str,
+            alert_body: str,
+            dedup: str | None = None,
+            severity: typing.Literal["critical", "error", "warning", "info"] | None = None,
+            event_type: typing.Literal["trigger", "acknowledge", "resolve"] | None = None,
+            source: str = "custom_event",
     ) -> typing.Dict[str, typing.Any]:
         """
         Builds the payload for an event alert.
@@ -383,13 +392,13 @@ def _build_alert(
         }
 
     def _send_alert(
-        self,
-        title: str,
-        body: str,
-        dedup: str | None = None,
-        severity: typing.Literal["critical", "error", "warning", "info"] | None = None,
-        event_type: typing.Literal["trigger", "acknowledge", "resolve"] | None = None,
-        source: str = "custom_event",
+            self,
+            title: str,
+            body: str,
+            dedup: str | None = None,
+            severity: typing.Literal["critical", "error", "warning", "info"] | None = None,
+            event_type: typing.Literal["trigger", "acknowledge", "resolve"] | None = None,
+            source: str = "custom_event",
     ):
         """
         Sends PagerDuty Alert
@@ -417,12 +426,12 @@ def _send_alert(
         return result.json()
 
     def _trigger_incident(
-        self,
-        service_id: str,
-        title: str,
-        body: dict,
-        requester: str,
-        incident_key: str | None = None,
+            self,
+            service_id: str,
+            title: str,
+            body: dict,
+            requester: str,
+            incident_key: str | None = None,
     ):
         """Triggers an incident via the V2 REST API using sample data."""
 
@@ -455,11 +464,11 @@ def dispose(self):
         pass
 
     def setup_incident_webhook(
-        self,
-        tenant_id: str,
-        keep_api_url: str,
-        api_key: str,
-        setup_alerts: bool = True,
+            self,
+            tenant_id: str,
+            keep_api_url: str,
+            api_key: str,
+            setup_alerts: bool = True,
     ):
         self.logger.info("Setting up Pagerduty webhook")
 
@@ -505,7 +514,14 @@ def setup_incident_webhook(
                     "incident.triggered",
                     "incident.unacknowledged",
                 ],
-                "filter": {"type": "account_reference"},
+                "filter": (
+                    {
+                        "type": "service_reference",
+                        "id": self.authentication_config.service_id,
+                    }
+                    if self.authentication_config.service_id
+                    else {"type": "account_reference"}
+                ),
             },
         }
         if webhook_exists:
@@ -530,17 +546,17 @@ def setup_incident_webhook(
         self.logger.info("Webhook created")
 
     def _notify(
-        self,
-        title: str = "",
-        alert_body: str = "",
-        dedup: str = "",
-        service_id: str = "",
-        requester: str = "",
-        incident_id: str = "",
-        event_type: typing.Literal["trigger", "acknowledge", "resolve"] | None = None,
-        severity: typing.Literal["critical", "error", "warning", "info"] | None = None,
-        source: str = "custom_event",
-        **kwargs: dict,
+            self,
+            title: str = "",
+            alert_body: str = "",
+            dedup: str = "",
+            service_id: str = "",
+            requester: str = "",
+            incident_id: str = "",
+            event_type: typing.Literal["trigger", "acknowledge", "resolve"] | None = None,
+            severity: typing.Literal["critical", "error", "warning", "info"] | None = None,
+            source: str = "custom_event",
+            **kwargs: dict,
     ):
         """
         Create a PagerDuty alert.
@@ -576,7 +592,7 @@ def _query(self, incident_id: str = None):
         )
 
     def _format_alert(
-        event: dict, provider_instance: "BaseProvider" = None
+            event: dict, provider_instance: "BaseProvider" = None
     ) -> AlertDto:
         # If somebody connected the provider before we refactored it
         old_format_event = event.get("event", {})
@@ -678,14 +694,17 @@ def __get_all_incidents_or_alerts(self, incident_id: str = None):
                     url += f"/{incident_id}/alerts"
                     include = ["teams", "services"]
                     resource = "alerts"
+                params = {
+                    "include[]": include,
+                    "offset": offset,
+                    "limit": 100,
+                }
+                if not incident_id and self.authentication_config.service_id:
+                    params["service_ids[]"] = [self.authentication_config.service_id]
                 response = requests.get(
                     url=url,
                     headers=self.__get_headers(),
-                    params={
-                        "include[]": include,
-                        "offset": offset,
-                        "limit": 100,
-                    },
+                    params=params,
                 )
                 response.raise_for_status()
                 response = response.json()
@@ -696,7 +715,7 @@ def __get_all_incidents_or_alerts(self, incident_id: str = None):
             paginated_response.extend(response.get(resource, []))
             self.logger.info("Fetched incidents or alerts", extra={"offset": offset})
             # No more results
-            if response.get("more", False) == False:
+            if not response.get("more", False):
                 self.logger.info("No more incidents or alerts")
                 break
         self.logger.info(
@@ -817,7 +836,7 @@ def _get_incident_id(incident_id: str) -> str:
 
     @staticmethod
     def _format_incident(
-        event: dict, provider_instance: "BaseProvider" = None
+            event: dict, provider_instance: "BaseProvider" = None
     ) -> IncidentDto | list[IncidentDto]:
 
         event = event["event"]["data"]