From 0b91ccce47bd821ed69ea2628f80c98f83981033 Mon Sep 17 00:00:00 2001
From: Erik Johnston <erikj@element.io>
Date: Tue, 14 May 2024 14:39:04 +0100
Subject: [PATCH] Improve perf of sync device lists (#17191)

It's almost always more efficient to query the rooms that have device
list changes, rather than looking at the list of all users whose devices
have changed and then look for shared rooms.
---
 changelog.d/17191.misc                    |  1 +
 synapse/handlers/sync.py                  | 37 ++++-------------------
 synapse/storage/databases/main/devices.py | 17 ++---------
 3 files changed, 9 insertions(+), 46 deletions(-)
 create mode 100644 changelog.d/17191.misc

diff --git a/changelog.d/17191.misc b/changelog.d/17191.misc
new file mode 100644
index 000000000..bd55eeaa3
--- /dev/null
+++ b/changelog.d/17191.misc
@@ -0,0 +1 @@
+Improve performance of calculating device lists changes in `/sync`.
diff --git a/synapse/handlers/sync.py b/synapse/handlers/sync.py
index 8ff45a335..0bef58351 100644
--- a/synapse/handlers/sync.py
+++ b/synapse/handlers/sync.py
@@ -1803,38 +1803,13 @@ class SyncHandler:
 
         # Step 1a, check for changes in devices of users we share a room
         # with
-        #
-        # We do this in two different ways depending on what we have cached.
-        # If we already have a list of all the user that have changed since
-        # the last sync then it's likely more efficient to compare the rooms
-        # they're in with the rooms the syncing user is in.
-        #
-        # If we don't have that info cached then we get all the users that
-        # share a room with our user and check if those users have changed.
-        cache_result = self.store.get_cached_device_list_changes(
-            since_token.device_list_key
-        )
-        if cache_result.hit:
-            changed_users = cache_result.entities
-
-            result = await self.store.get_rooms_for_users(changed_users)
-
-            for changed_user_id, entries in result.items():
-                # Check if the changed user shares any rooms with the user,
-                # or if the changed user is the syncing user (as we always
-                # want to include device list updates of their own devices).
-                if user_id == changed_user_id or any(
-                    rid in joined_rooms for rid in entries
-                ):
-                    users_that_have_changed.add(changed_user_id)
-        else:
-            users_that_have_changed = (
-                await self._device_handler.get_device_changes_in_shared_rooms(
-                    user_id,
-                    sync_result_builder.joined_room_ids,
-                    from_token=since_token,
-                )
+        users_that_have_changed = (
+            await self._device_handler.get_device_changes_in_shared_rooms(
+                user_id,
+                sync_result_builder.joined_room_ids,
+                from_token=since_token,
             )
+        )
 
         # Step 1b, check for newly joined rooms
         for room_id in newly_joined_rooms:
diff --git a/synapse/storage/databases/main/devices.py b/synapse/storage/databases/main/devices.py
index 8dbcb3f5a..d98f0593b 100644
--- a/synapse/storage/databases/main/devices.py
+++ b/synapse/storage/databases/main/devices.py
@@ -70,10 +70,7 @@ from synapse.types import (
 from synapse.util import json_decoder, json_encoder
 from synapse.util.caches.descriptors import cached, cachedList
 from synapse.util.caches.lrucache import LruCache
-from synapse.util.caches.stream_change_cache import (
-    AllEntitiesChangedResult,
-    StreamChangeCache,
-)
+from synapse.util.caches.stream_change_cache import StreamChangeCache
 from synapse.util.cancellation import cancellable
 from synapse.util.iterutils import batch_iter
 from synapse.util.stringutils import shortstr
@@ -832,16 +829,6 @@ class DeviceWorkerStore(RoomMemberWorkerStore, EndToEndKeyWorkerStore):
         )
         return {device[0]: db_to_json(device[1]) for device in devices}
 
-    def get_cached_device_list_changes(
-        self,
-        from_key: int,
-    ) -> AllEntitiesChangedResult:
-        """Get set of users whose devices have changed since `from_key`, or None
-        if that information is not in our cache.
-        """
-
-        return self._device_list_stream_cache.get_all_entities_changed(from_key)
-
     @cancellable
     async def get_all_devices_changed(
         self,
@@ -1475,7 +1462,7 @@ class DeviceWorkerStore(RoomMemberWorkerStore, EndToEndKeyWorkerStore):
 
         sql = """
             SELECT DISTINCT user_id FROM device_lists_changes_in_room
-            WHERE {clause} AND stream_id >= ?
+            WHERE {clause} AND stream_id > ?
         """
 
         def _get_device_list_changes_in_rooms_txn(