fix(persons): limit to maximum of 2500 distinct_ids for cross-db join

posthog/hogql_queries/events_query_runner.py

@@ -33,6 +33,8 @@
     "created_at",
 ]
 
+MAX_LIMIT_DISTINCT_IDS = 2500
+
 
 class EventsQueryRunner(QueryRunner):
     query: EventsQuery
@@ -116,7 +118,7 @@ def to_query(self) -> ast.SelectQuery:
                 if self.query.personId:
                     with self.timings.measure("person_id"):
                         person: Optional[Person] = get_pk_or_uuid(
-                            Person.objects.filter(team=self.team), self.query.personId
+                            Person.objects.filter(team=self.team)[:MAX_LIMIT_DISTINCT_IDS], self.query.personId
                         ).first()
                         distinct_ids = person.distinct_ids if person is not None else []
                         ids_list = list(map(str, distinct_ids))

posthog/models/event/query_event_list.py

@@ -9,6 +9,7 @@
 from posthog.clickhouse.client.connection import Workload
 from posthog.hogql.constants import DEFAULT_RETURNED_ROWS
 from posthog.hogql.context import HogQLContext
+from posthog.hogql_queries.events_query_runner import MAX_LIMIT_DISTINCT_IDS
 from posthog.models import Action, Filter, Person, Team
 from posthog.models.action.util import format_action_filter
 from posthog.models.event.sql import (
@@ -44,7 +45,7 @@ def determine_event_conditions(
             params.update({"before": timestamp})
         elif k == "person_id":
             result += """AND distinct_id IN (%(distinct_ids)s) """
-            person = get_pk_or_uuid(Person.objects.filter(team=team), v).first()
+            person = get_pk_or_uuid(Person.objects.filter(team=team)[:MAX_LIMIT_DISTINCT_IDS], v).first()
             distinct_ids = person.distinct_ids if person is not None else []
             params.update({"distinct_ids": list(map(str, distinct_ids))})
         elif k == "distinct_id":