Use `execute_values` more in PostgreSQL (#10754)

`execute_values` is a faster version of `execute_batch`.

2 files changed, 43 insertions, 19 deletions

diff --git a/changelog.d/10754.misc b/changelog.d/10754.misc
new file mode 100644
index 0000000000..3b7acff03f
--- /dev/null
+++ b/changelog.d/10754.misc
@@ -0,0 +1 @@
+Minor speed ups when joining large rooms over federation.
diff --git a/synapse/storage/database.py b/synapse/storage/database.py
index 95d2caff62..0084d9f96c 100644
--- a/synapse/storage/database.py
+++ b/synapse/storage/database.py
@@ -280,18 +280,18 @@ class LoggingTransaction:
         else:
             self.executemany(sql, args)
 
-    def execute_values(self, sql: str, *args: Any) -> List[Tuple]:
+    def execute_values(self, sql: str, *args: Any, fetch: bool = True) -> List[Tuple]:
         """Corresponds to psycopg2.extras.execute_values. Only available when
         using postgres.
 
-        Always sets fetch=True when caling `execute_values`, so will return the
-        results.
+        The `fetch` parameter must be set to False if the query does not return
+        rows (e.g. INSERTs).
         """
         assert isinstance(self.database_engine, PostgresEngine)
         from psycopg2.extras import execute_values  # type: ignore
 
         return self._do_execute(
-            lambda *x: execute_values(self.txn, *x, fetch=True), sql, *args
+            lambda *x: execute_values(self.txn, *x, fetch=fetch), sql, *args
         )
 
     def execute(self, sql: str, *args: Any) -> None:
@@ -920,13 +920,23 @@ class DatabasePool:
             if k != keys[0]:
                 raise RuntimeError("All items must have the same keys")
 
-        sql = "INSERT INTO %s (%s) VALUES(%s)" % (
-            table,
-            ", ".join(k for k in keys[0]),
-            ", ".join("?" for _ in keys[0]),
-        )
+        if isinstance(txn.database_engine, PostgresEngine):
+            # We use `execute_values` as it can be a lot faster than `execute_batch`,
+            # but it's only available on postgres.
+            sql = "INSERT INTO %s (%s) VALUES ?" % (
+                table,
+                ", ".join(k for k in keys[0]),
+            )
 
-        txn.execute_batch(sql, vals)
+            txn.execute_values(sql, vals, fetch=False)
+        else:
+            sql = "INSERT INTO %s (%s) VALUES(%s)" % (
+                table,
+                ", ".join(k for k in keys[0]),
+                ", ".join("?" for _ in keys[0]),
+            )
+
+            txn.execute_batch(sql, vals)
 
     async def simple_upsert(
         self,
@@ -1281,20 +1291,33 @@ class DatabasePool:
                 k + "=EXCLUDED." + k for k in value_names
             )
 
-        sql = "INSERT INTO %s (%s) VALUES (%s) ON CONFLICT (%s) DO %s" % (
-            table,
-            ", ".join(k for k in allnames),
-            ", ".join("?" for _ in allnames),
-            ", ".join(key_names),
-            latter,
-        )
-
         args = []
 
         for x, y in zip(key_values, value_values):
             args.append(tuple(x) + tuple(y))
 
-        return txn.execute_batch(sql, args)
+        if isinstance(txn.database_engine, PostgresEngine):
+            # We use `execute_values` as it can be a lot faster than `execute_batch`,
+            # but it's only available on postgres.
+            sql = "INSERT INTO %s (%s) VALUES ? ON CONFLICT (%s) DO %s" % (
+                table,
+                ", ".join(k for k in allnames),
+                ", ".join(key_names),
+                latter,
+            )
+
+            txn.execute_values(sql, args, fetch=False)
+
+        else:
+            sql = "INSERT INTO %s (%s) VALUES (%s) ON CONFLICT (%s) DO %s" % (
+                table,
+                ", ".join(k for k in allnames),
+                ", ".join("?" for _ in allnames),
+                ", ".join(key_names),
+                latter,
+            )
+
+            return txn.execute_batch(sql, args)
 
     @overload
     async def simple_select_one(