ref(grouping): Increase config transition and move config tests (#112508)

lobsterkatie · george-sentry · commit 09abcc3f8f8a · 2026-04-09T14:13:00.000-07:00
When a project's grouping config is updated, we put the project in a transition period, during which we calculate the old hash if we don't find the new hash, so we can attach the new hash to the old hash's group (assuming the old hash exists). This prevents grouping configs from creating new groups even though event data hasn't changed.

It used to be that during the transition period, we _always_ calculated the old hash in addition to the new hash, even if we'd already linked the new hash to an existing group (if any). Since the vast majority of events come in to existing groups, this meant a ton of wasted calculation, and an unacceptable slow-down during ingest if too many projects were in transition at once. To limit the number of projects transitioning at any one time, we set the transition period to only last a week.

Now, however, we limit the double calculation to the first time we see an event from a given group, rather than every time we see an event from that group. Because this theoretically reduced (by multiple orders of magnitude) the extra work a transition would entail, once we made that change, we increased the transition period to a month. We still hadn't seen it in action, though.

That changed a few months ago, when we introduced a new grouping config. Sure enough, though there was an initial spike of double calculations (at about 7% of events), it dropped extremely quickly, and within a few hours it was down under 1% of events (and has only continued to fall since). Even more encouragingly, there was no noticeable change to average ingest time.

Now that we have that assurance, we can increase the transition period to the full 90 days it takes for groups to age out, thus preventing even the few groups which go more than a month between events from being recreated. This PR does just that, by changing the `SENTRY_GROUPING_CONFIG_TRANSITION_DURATION` constant. While we're dealing with grouping config stuff, it also moves the relevant tests to their own file, continuing the effort to have `test_event_manager_grouping.py` not be literally a thousand lines long.
diff --git a/src/sentry/conf/server.py b/src/sentry/conf/server.py
@@ -2817,7 +2817,7 @@ def custom_parameter_sort(parameter: dict) -> tuple[str, int]:
 BETA_GROUPING_CONFIG = ""
 
 # How long the migration phase for grouping lasts
-SENTRY_GROUPING_CONFIG_TRANSITION_DURATION = 30 * 24 * 3600  # 30 days
+SENTRY_GROUPING_CONFIG_TRANSITION_DURATION = 90 * 24 * 3600  # 90 days, until groups age out
 
 SENTRY_USE_GRANIAN = True
 
diff --git a/tests/sentry/event_manager/test_event_manager_grouping.py b/tests/sentry/event_manager/test_event_manager_grouping.py
@@ -6,40 +6,34 @@
 from time import time
 from typing import Any
 from unittest import mock
-from unittest.mock import ANY, MagicMock, patch
+from unittest.mock import MagicMock, patch
 
 import pytest
 from django.core.cache import cache
 
-from sentry import audit_log
-from sentry.conf.server import DEFAULT_GROUPING_CONFIG, SENTRY_GROUPING_CONFIG_TRANSITION_DURATION
+from sentry.conf.server import DEFAULT_GROUPING_CONFIG
 from sentry.event_manager import _get_updated_group_title
 from sentry.eventtypes.base import DefaultEvent
 from sentry.exceptions import HashDiscarded
-from sentry.grouping.api import get_grouping_config_dict_for_project
 from sentry.grouping.ingest.caching import (
     get_grouphash_existence_cache_key,
     get_grouphash_object_cache_key,
 )
-from sentry.grouping.ingest.config import update_or_set_grouping_config_if_needed
 from sentry.grouping.ingest.hashing import (
     GROUPHASH_CACHE_EXPIRY_SECONDS,
     find_grouphash_with_group,
     get_or_create_grouphashes,
 )
-from sentry.models.auditlogentry import AuditLogEntry
 from sentry.models.group import Group
 from sentry.models.grouphash import GroupHash
 from sentry.models.grouptombstone import GroupTombstone
-from sentry.models.options.project_option import ProjectOption
 from sentry.models.project import Project
 from sentry.services.eventstore.models import Event
 from sentry.testutils.cases import TestCase
 from sentry.testutils.helpers.eventprocessing import save_new_event
 from sentry.testutils.helpers.options import override_options
 from sentry.testutils.pytest.fixtures import django_db_all
 from sentry.testutils.pytest.mocking import count_matching_calls
-from sentry.testutils.silo import assume_test_silo_mode_of
 from sentry.testutils.skips import requires_snuba
 from tests.sentry.grouping import NO_MSG_PARAM_CONFIG
 
@@ -186,186 +180,6 @@ def test_updates_group_metadata(self) -> None:
         assert group.message == event2.message
         assert group.data["metadata"]["title"] == event2.title
 
-    def test_loads_default_config_if_stored_config_option_is_invalid(self) -> None:
-        self.project.update_option("sentry:grouping_config", "dogs.are.great")
-        config_dict = get_grouping_config_dict_for_project(self.project)
-        assert config_dict["id"] == DEFAULT_GROUPING_CONFIG
-
-        self.project.update_option("sentry:grouping_config", {"not": "a string"})
-        config_dict = get_grouping_config_dict_for_project(self.project)
-        assert config_dict["id"] == DEFAULT_GROUPING_CONFIG
-
-    def test_auto_updates_grouping_config_even_if_config_is_gone(self) -> None:
-        """This tests that setups with deprecated configs will auto-upgrade."""
-        self.project.update_option("sentry:grouping_config", "non_existing_config")
-        save_new_event({"message": "foo"}, self.project)
-        assert self.project.get_option("sentry:grouping_config") == DEFAULT_GROUPING_CONFIG
-        assert self.project.get_option("sentry:secondary_grouping_config") is None
-
-    def test_auto_updates_grouping_config(self) -> None:
-        self.project.update_option("sentry:grouping_config", NO_MSG_PARAM_CONFIG)
-        # Set platform to prevent additional audit log entry from platform inference
-        self.project.platform = "python"
-        self.project.save()
-
-        save_new_event({"message": "Adopt don't shop"}, self.project)
-        assert self.project.get_option("sentry:grouping_config") == DEFAULT_GROUPING_CONFIG
-
-        with assume_test_silo_mode_of(AuditLogEntry):
-            audit_log_entry = AuditLogEntry.objects.get()
-
-        assert audit_log_entry.event == audit_log.get_event_id("PROJECT_EDIT")
-        assert audit_log_entry.actor_label == "Sentry"
-
-        assert audit_log_entry.data == {
-            "sentry:grouping_config": DEFAULT_GROUPING_CONFIG,
-            "sentry:secondary_grouping_config": NO_MSG_PARAM_CONFIG,
-            "sentry:secondary_grouping_expiry": ANY,  # tested separately below
-            "id": self.project.id,
-            "slug": self.project.slug,
-            "name": self.project.name,
-            "status": 0,
-            "public": False,
-        }
-
-        # When the config upgrade is actually happening, the expiry value is set before the
-        # audit log entry is created, which means the expiry is based on a timestamp
-        # ever-so-slightly before the audit log entry's timestamp, making a one-second tolerance
-        # necessary.
-        actual_expiry = audit_log_entry.data["sentry:secondary_grouping_expiry"]
-        expected_expiry = (
-            int(audit_log_entry.datetime.timestamp()) + SENTRY_GROUPING_CONFIG_TRANSITION_DURATION
-        )
-        assert actual_expiry == expected_expiry or actual_expiry == expected_expiry - 1
-
-    @patch(
-        "sentry.event_manager.update_or_set_grouping_config_if_needed",
-        wraps=update_or_set_grouping_config_if_needed,
-    )
-    def test_sets_default_grouping_config_project_option_if_missing(
-        self, update_config_spy: MagicMock
-    ):
-        # To start, the project defaults to the current config but doesn't have its own config
-        # option set in the DB
-        assert self.project.get_option("sentry:grouping_config") == DEFAULT_GROUPING_CONFIG
-        assert (
-            ProjectOption.objects.filter(
-                project_id=self.project.id, key="sentry:grouping_config"
-            ).first()
-            is None
-        )
-
-        save_new_event({"message": "Dogs are great!"}, self.project)
-
-        update_config_spy.assert_called_with(self.project, "ingest")
-
-        # After the function has been called, the config still defaults to the current one (and no
-        # transition has started), but the project now has its own config record in the DB
-        assert self.project.get_option("sentry:grouping_config") == DEFAULT_GROUPING_CONFIG
-        assert self.project.get_option("sentry:secondary_grouping_config") is None
-        assert self.project.get_option("sentry:secondary_grouping_expiry") == 0
-        assert ProjectOption.objects.filter(
-            project_id=self.project.id, key="sentry:grouping_config"
-        ).exists()
-
-    @patch(
-        "sentry.event_manager.update_or_set_grouping_config_if_needed",
-        wraps=update_or_set_grouping_config_if_needed,
-    )
-    def test_no_ops_if_grouping_config_project_option_exists_and_is_current(
-        self, update_config_spy: MagicMock
-    ):
-        self.project.update_option("sentry:grouping_config", DEFAULT_GROUPING_CONFIG)
-
-        assert self.project.get_option("sentry:grouping_config") == DEFAULT_GROUPING_CONFIG
-        assert ProjectOption.objects.filter(
-            project_id=self.project.id, key="sentry:grouping_config"
-        ).exists()
-
-        save_new_event({"message": "Dogs are great!"}, self.project)
-
-        update_config_spy.assert_called_with(self.project, "ingest")
-
-        # After the function has been called, the config still defaults to the current one and no
-        # transition has started
-        assert self.project.get_option("sentry:grouping_config") == DEFAULT_GROUPING_CONFIG
-        assert self.project.get_option("sentry:secondary_grouping_config") is None
-        assert self.project.get_option("sentry:secondary_grouping_expiry") == 0
-
-    @patch(
-        "sentry.event_manager.update_or_set_grouping_config_if_needed",
-        wraps=update_or_set_grouping_config_if_needed,
-    )
-    def test_no_ops_if_sample_rate_test_fails(self, update_config_spy: MagicMock):
-        with (
-            # Ensure our die roll will fall outside the sample rate
-            patch("sentry.grouping.ingest.config.random", return_value=0.1121),
-            override_options({"grouping.config_transition.config_upgrade_sample_rate": 0.0908}),
-        ):
-            self.project.update_option("sentry:grouping_config", NO_MSG_PARAM_CONFIG)
-            assert self.project.get_option("sentry:grouping_config") == NO_MSG_PARAM_CONFIG
-
-            save_new_event({"message": "Dogs are great!"}, self.project)
-
-            update_config_spy.assert_called_with(self.project, "ingest")
-
-            # After the function has been called, the config hasn't changed and no transition has
-            # started
-            assert self.project.get_option("sentry:grouping_config") == NO_MSG_PARAM_CONFIG
-            assert self.project.get_option("sentry:secondary_grouping_config") is None
-            assert self.project.get_option("sentry:secondary_grouping_expiry") == 0
-
-    @patch(
-        "sentry.event_manager.update_or_set_grouping_config_if_needed",
-        wraps=update_or_set_grouping_config_if_needed,
-    )
-    def test_ignores_sample_rate_if_current_config_is_invalid(self, update_config_spy: MagicMock):
-        with (
-            # Ensure our die roll will fall outside the sample rate
-            patch("sentry.grouping.ingest.config.random", return_value=0.1121),
-            override_options({"grouping.config_transition.config_upgrade_sample_rate": 0.0908}),
-        ):
-            self.project.update_option("sentry:grouping_config", "not_a_real_config")
-            assert self.project.get_option("sentry:grouping_config") == "not_a_real_config"
-
-            save_new_event({"message": "Dogs are great!"}, self.project)
-
-            update_config_spy.assert_called_with(self.project, "ingest")
-
-            # The config has been updated, but no transition has started because we can't calculate
-            # a secondary hash using a config that doesn't exist
-            assert self.project.get_option("sentry:grouping_config") == DEFAULT_GROUPING_CONFIG
-            assert self.project.get_option("sentry:secondary_grouping_config") is None
-            assert self.project.get_option("sentry:secondary_grouping_expiry") == 0
-
-    @patch(
-        "sentry.event_manager.update_or_set_grouping_config_if_needed",
-        wraps=update_or_set_grouping_config_if_needed,
-    )
-    def test_ignores_sample_rate_if_no_record_exists(self, update_config_spy: MagicMock):
-        with (
-            # Ensure our die roll will fall outside the sample rate
-            patch("sentry.grouping.ingest.config.random", return_value=0.1121),
-            override_options({"grouping.config_transition.config_upgrade_sample_rate": 0.0908}),
-        ):
-            assert self.project.get_option("sentry:grouping_config") == DEFAULT_GROUPING_CONFIG
-            assert not ProjectOption.objects.filter(
-                project_id=self.project.id, key="sentry:grouping_config"
-            ).exists()
-
-            save_new_event({"message": "Dogs are great!"}, self.project)
-
-            update_config_spy.assert_called_with(self.project, "ingest")
-
-            # The config hasn't been updated, but now the project has its own record. No transition
-            # has started because the config was already up to date.
-            assert self.project.get_option("sentry:grouping_config") == DEFAULT_GROUPING_CONFIG
-            assert ProjectOption.objects.filter(
-                project_id=self.project.id, key="sentry:grouping_config"
-            ).exists()
-            assert self.project.get_option("sentry:secondary_grouping_config") is None
-            assert self.project.get_option("sentry:secondary_grouping_expiry") == 0
-
 
 class GroupHashCachingTest(TestCase):
     @contextmanager
diff --git a/tests/sentry/grouping/test_config.py b/tests/sentry/grouping/test_config.py
