Minor query optimizations (#4325)

Related to some API latency issues we were noticing.

engine/apps/alerts/models/alert_group.py

@@ -511,17 +511,8 @@ class AlertGroup(AlertGroupSlackRenderingMixin, EscalationSnapshotMixin, models.
 
     @property
     def telegram_permalink(self) -> typing.Optional[str]:
-        """
-        This property will attempt to access an attribute, `prefetched_telegram_messages`, representing a list of
-        prefetched telegram messages. If this attribute does not exist, it falls back to performing a query.
-
-        See `apps.public_api.serializers.incidents.IncidentSerializer.PREFETCH_RELATED` as an example.
-        """
         from apps.telegram.models.message import TelegramMessage
 
-        if hasattr(self, "prefetched_telegram_messages"):
-            return self.prefetched_telegram_messages[0].link if self.prefetched_telegram_messages else None
-
         main_telegram_message = self.telegram_messages.filter(
             chat_id__startswith="-", message_type=TelegramMessage.ALERT_GROUP_MESSAGE
         ).first()

engine/apps/public_api/serializers/incidents.py

@@ -1,8 +1,6 @@
-from django.db.models import Prefetch
 from rest_framework import serializers
 
 from apps.alerts.models import AlertGroup
-from apps.telegram.models.message import TelegramMessage
 from common.api_helpers.custom_fields import UserIdField
 from common.api_helpers.mixins import EagerLoadingMixin
 
@@ -19,14 +17,6 @@ class IncidentSerializer(EagerLoadingMixin, serializers.ModelSerializer):
     resolved_by = UserIdField(read_only=True, source="resolved_by_user")
 
     SELECT_RELATED = ["channel", "channel_filter", "slack_message", "channel__organization"]
-    PREFETCH_RELATED = [
-        "alerts",
-        Prefetch(
-            "telegram_messages",
-            TelegramMessage.objects.filter(chat_id__startswith="-", message_type=TelegramMessage.ALERT_GROUP_MESSAGE),
-            to_attr="prefetched_telegram_messages",
-        ),
-    ]
 
     class Meta:
         model = AlertGroup
@@ -50,7 +40,7 @@ class IncidentSerializer(EagerLoadingMixin, serializers.ModelSerializer):
         return obj.web_title_cache
 
     def get_alerts_count(self, obj):
-        return len(obj.alerts.all())
+        return obj.alerts.count()
 
     def get_state(self, obj):
         return obj.state