From 76f6b87949d613f4397bd20cc744384ff1477775 Mon Sep 17 00:00:00 2001
From: Erik Johnston <erik@matrix.org>
Date: Wed, 8 Sep 2021 16:38:50 +0100
Subject: [PATCH 1/2] Don't needlessly batch in `add_event_to_cache`

We've already batched up the events previously, and assume in other
places in the events.py file that we have. Removing this makes it easier
to adjust the batch sizes in one place.
---
 synapse/storage/databases/main/events.py | 51 +++++++++++-------------
 1 file changed, 24 insertions(+), 27 deletions(-)

diff --git a/synapse/storage/databases/main/events.py b/synapse/storage/databases/main/events.py
index 14ada8a8b359..8e691678e543 100644
--- a/synapse/storage/databases/main/events.py
+++ b/synapse/storage/databases/main/events.py
@@ -1547,35 +1547,32 @@ def _add_to_cache(self, txn, events_and_contexts):
         to_prefill = []
 
         rows = []
-        N = 200
-        for i in range(0, len(events_and_contexts), N):
-            ev_map = {e[0].event_id: e[0] for e in events_and_contexts[i : i + N]}
-            if not ev_map:
-                break
-
-            sql = (
-                "SELECT "
-                " e.event_id as event_id, "
-                " r.redacts as redacts,"
-                " rej.event_id as rejects "
-                " FROM events as e"
-                " LEFT JOIN rejections as rej USING (event_id)"
-                " LEFT JOIN redactions as r ON e.event_id = r.redacts"
-                " WHERE "
-            )
 
-            clause, args = make_in_list_sql_clause(
-                self.database_engine, "e.event_id", list(ev_map)
-            )
+        ev_map = {e.event_id: e for e, _ in events_and_contexts}
+        if not ev_map:
+            return
 
-            txn.execute(sql + clause, args)
-            rows = self.db_pool.cursor_to_dict(txn)
-            for row in rows:
-                event = ev_map[row["event_id"]]
-                if not row["rejects"] and not row["redacts"]:
-                    to_prefill.append(
-                        _EventCacheEntry(event=event, redacted_event=None)
-                    )
+        sql = (
+            "SELECT "
+            " e.event_id as event_id, "
+            " r.redacts as redacts,"
+            " rej.event_id as rejects "
+            " FROM events as e"
+            " LEFT JOIN rejections as rej USING (event_id)"
+            " LEFT JOIN redactions as r ON e.event_id = r.redacts"
+            " WHERE "
+        )
+
+        clause, args = make_in_list_sql_clause(
+            self.database_engine, "e.event_id", list(ev_map)
+        )
+
+        txn.execute(sql + clause, args)
+        rows = self.db_pool.cursor_to_dict(txn)
+        for row in rows:
+            event = ev_map[row["event_id"]]
+            if not row["rejects"] and not row["redacts"]:
+                to_prefill.append(_EventCacheEntry(event=event, redacted_event=None))
 
         def prefill():
             for cache_entry in to_prefill:

From cd38f9fa2a92a5c0ba1522f88774347e6fd4881a Mon Sep 17 00:00:00 2001
From: Erik Johnston <erik@matrix.org>
Date: Wed, 8 Sep 2021 16:40:40 +0100
Subject: [PATCH 2/2] Newsfile

---
 changelog.d/10784.misc | 1 +
 1 file changed, 1 insertion(+)
 create mode 100644 changelog.d/10784.misc

diff --git a/changelog.d/10784.misc b/changelog.d/10784.misc
new file mode 100644
index 000000000000..3b7acff03f4f
--- /dev/null
+++ b/changelog.d/10784.misc
@@ -0,0 +1 @@
+Minor speed ups when joining large rooms over federation.