From 922788c6043138165c025c78effeda87de842bab Mon Sep 17 00:00:00 2001
From: Patrick Cloke <clokep@users.noreply.github.com>
Date: Wed, 3 Mar 2021 11:04:08 -0500
Subject: [PATCH] Purge chain cover tables when purging events. (#9498)

---
 changelog.d/9498.bugfix                       |  1 +
 .../storage/databases/main/purge_events.py    | 42 ++++++++++++++++---
 synapse/storage/purge_events.py               |  5 ---
 3 files changed, 38 insertions(+), 10 deletions(-)
 create mode 100644 changelog.d/9498.bugfix

diff --git a/changelog.d/9498.bugfix b/changelog.d/9498.bugfix
new file mode 100644
index 000000000..dce0ad092
--- /dev/null
+++ b/changelog.d/9498.bugfix
@@ -0,0 +1 @@
+Properly purge the event chain cover index when purging history.
diff --git a/synapse/storage/databases/main/purge_events.py b/synapse/storage/databases/main/purge_events.py
index ecfc9f20b..0836e4af4 100644
--- a/synapse/storage/databases/main/purge_events.py
+++ b/synapse/storage/databases/main/purge_events.py
@@ -28,7 +28,10 @@ class PurgeEventsStore(StateGroupWorkerStore, SQLBaseStore):
     async def purge_history(
         self, room_id: str, token: str, delete_local_events: bool
     ) -> Set[int]:
-        """Deletes room history before a certain point
+        """Deletes room history before a certain point.
+
+        Note that only a single purge can occur at once, this is guaranteed via
+        a higher level (in the PaginationHandler).
 
         Args:
             room_id:
@@ -52,7 +55,9 @@ class PurgeEventsStore(StateGroupWorkerStore, SQLBaseStore):
             delete_local_events,
         )
 
-    def _purge_history_txn(self, txn, room_id, token, delete_local_events):
+    def _purge_history_txn(
+        self, txn, room_id: str, token: RoomStreamToken, delete_local_events: bool
+    ) -> Set[int]:
         # Tables that should be pruned:
         #     event_auth
         #     event_backward_extremities
@@ -103,7 +108,7 @@ class PurgeEventsStore(StateGroupWorkerStore, SQLBaseStore):
         if max_depth < token.topological:
             # We need to ensure we don't delete all the events from the database
             # otherwise we wouldn't be able to send any events (due to not
-            # having any backwards extremeties)
+            # having any backwards extremities)
             raise SynapseError(
                 400, "topological_ordering is greater than forward extremeties"
             )
@@ -154,7 +159,7 @@ class PurgeEventsStore(StateGroupWorkerStore, SQLBaseStore):
 
         logger.info("[purge] Finding new backward extremities")
 
-        # We calculate the new entries for the backward extremeties by finding
+        # We calculate the new entries for the backward extremities by finding
         # events to be purged that are pointed to by events we're not going to
         # purge.
         txn.execute(
@@ -296,7 +301,7 @@ class PurgeEventsStore(StateGroupWorkerStore, SQLBaseStore):
             "purge_room", self._purge_room_txn, room_id
         )
 
-    def _purge_room_txn(self, txn, room_id):
+    def _purge_room_txn(self, txn, room_id: str) -> List[int]:
         # First we fetch all the state groups that should be deleted, before
         # we delete that information.
         txn.execute(
@@ -310,6 +315,31 @@ class PurgeEventsStore(StateGroupWorkerStore, SQLBaseStore):
 
         state_groups = [row[0] for row in txn]
 
+        # Get all the auth chains that are referenced by events that are to be
+        # deleted.
+        txn.execute(
+            """
+            SELECT chain_id, sequence_number FROM events
+            LEFT JOIN event_auth_chains USING (event_id)
+            WHERE room_id = ?
+            """,
+            (room_id,),
+        )
+        referenced_chain_id_tuples = list(txn)
+
+        logger.info("[purge] removing events from event_auth_chain_links")
+        txn.executemany(
+            """
+            DELETE FROM event_auth_chain_links WHERE
+            (origin_chain_id = ? AND origin_sequence_number = ?) OR
+            (target_chain_id = ? AND target_sequence_number = ?)
+            """,
+            (
+                (chain_id, seq_num, chain_id, seq_num)
+                for (chain_id, seq_num) in referenced_chain_id_tuples
+            ),
+        )
+
         # Now we delete tables which lack an index on room_id but have one on event_id
         for table in (
             "event_auth",
@@ -319,6 +349,8 @@ class PurgeEventsStore(StateGroupWorkerStore, SQLBaseStore):
             "event_reference_hashes",
             "event_relations",
             "event_to_state_groups",
+            "event_auth_chains",
+            "event_auth_chain_to_calculate",
             "redactions",
             "rejections",
             "state_events",
diff --git a/synapse/storage/purge_events.py b/synapse/storage/purge_events.py
index 3c4908865..4dcd848c5 100644
--- a/synapse/storage/purge_events.py
+++ b/synapse/storage/purge_events.py
@@ -73,9 +73,6 @@ class PurgeEventsStorage:
         Returns:
             The set of state groups that can be deleted.
         """
-        # Graph of state group -> previous group
-        graph = {}
-
         # Set of events that we have found to be referenced by events
         referenced_groups = set()
 
@@ -111,8 +108,6 @@ class PurgeEventsStorage:
             next_to_search |= prevs
             state_groups_seen |= prevs
 
-            graph.update(edges)
-
         to_delete = state_groups_seen - referenced_groups
 
         return to_delete