Use a chain cover index to efficiently calculate auth chain difference (#8868)

1 files changed, 49 insertions, 2 deletions

diff --git a/synapse/storage/databases/main/room.py b/synapse/storage/databases/main/room.py
index 4650d0689b..284f2ce77c 100644
--- a/synapse/storage/databases/main/room.py
+++ b/synapse/storage/databases/main/room.py
@@ -84,7 +84,7 @@ class RoomWorkerStore(SQLBaseStore):
         return await self.db_pool.simple_select_one(
             table="rooms",
             keyvalues={"room_id": room_id},
-            retcols=("room_id", "is_public", "creator"),
+            retcols=("room_id", "is_public", "creator", "has_auth_chain_index"),
             desc="get_room",
             allow_none=True,
         )
@@ -1166,6 +1166,37 @@ class RoomBackgroundUpdateStore(SQLBaseStore):
         # It's overridden by RoomStore for the synapse master.
         raise NotImplementedError()
 
+    async def has_auth_chain_index(self, room_id: str) -> bool:
+        """Check if the room has (or can have) a chain cover index.
+
+        Defaults to True if we don't have an entry in `rooms` table nor any
+        events for the room.
+        """
+
+        has_auth_chain_index = await self.db_pool.simple_select_one_onecol(
+            table="rooms",
+            keyvalues={"room_id": room_id},
+            retcol="has_auth_chain_index",
+            desc="has_auth_chain_index",
+            allow_none=True,
+        )
+
+        if has_auth_chain_index:
+            return True
+
+        # It's possible that we already have events for the room in our DB
+        # without a corresponding room entry. If we do then we don't want to
+        # mark the room as having an auth chain cover index.
+        max_ordering = await self.db_pool.simple_select_one_onecol(
+            table="events",
+            keyvalues={"room_id": room_id},
+            retcol="MAX(stream_ordering)",
+            allow_none=True,
+            desc="upsert_room_on_join",
+        )
+
+        return max_ordering is None
+
 
 class RoomStore(RoomBackgroundUpdateStore, RoomWorkerStore, SearchStore):
     def __init__(self, database: DatabasePool, db_conn, hs):
@@ -1179,12 +1210,21 @@ class RoomStore(RoomBackgroundUpdateStore, RoomWorkerStore, SearchStore):
         Called when we join a room over federation, and overwrites any room version
         currently in the table.
         """
+        # It's possible that we already have events for the room in our DB
+        # without a corresponding room entry. If we do then we don't want to
+        # mark the room as having an auth chain cover index.
+        has_auth_chain_index = await self.has_auth_chain_index(room_id)
+
         await self.db_pool.simple_upsert(
             desc="upsert_room_on_join",
             table="rooms",
             keyvalues={"room_id": room_id},
             values={"room_version": room_version.identifier},
-            insertion_values={"is_public": False, "creator": ""},
+            insertion_values={
+                "is_public": False,
+                "creator": "",
+                "has_auth_chain_index": has_auth_chain_index,
+            },
             # rooms has a unique constraint on room_id, so no need to lock when doing an
             # emulated upsert.
             lock=False,
@@ -1219,6 +1259,7 @@ class RoomStore(RoomBackgroundUpdateStore, RoomWorkerStore, SearchStore):
                         "creator": room_creator_user_id,
                         "is_public": is_public,
                         "room_version": room_version.identifier,
+                        "has_auth_chain_index": True,
                     },
                 )
                 if is_public:
@@ -1247,6 +1288,11 @@ class RoomStore(RoomBackgroundUpdateStore, RoomWorkerStore, SearchStore):
         When we receive an invite or any other event over federation that may relate to a room
         we are not in, store the version of the room if we don't already know the room version.
         """
+        # It's possible that we already have events for the room in our DB
+        # without a corresponding room entry. If we do then we don't want to
+        # mark the room as having an auth chain cover index.
+        has_auth_chain_index = await self.has_auth_chain_index(room_id)
+
         await self.db_pool.simple_upsert(
             desc="maybe_store_room_on_outlier_membership",
             table="rooms",
@@ -1256,6 +1302,7 @@ class RoomStore(RoomBackgroundUpdateStore, RoomWorkerStore, SearchStore):
                 "room_version": room_version.identifier,
                 "is_public": False,
                 "creator": "",
+                "has_auth_chain_index": has_auth_chain_index,
             },
             # rooms has a unique constraint on room_id, so no need to lock when doing an
             # emulated upsert.