Use autocommit mode for single statement DB functions. (#8542)

Autocommit means that we don't wrap the functions in transactions, and instead get executed directly. Introduced in #8456. This will help:

1. reduce the number of `could not serialize access due to concurrent delete` errors that we see (though there are a few functions that often cause serialization errors that we don't fix here);
2. improve the DB performance, as it no longer needs to deal with the overhead of `REPEATABLE READ` isolation levels; and
3. improve wall clock speed of these functions, as we no longer need to send `BEGIN` and `COMMIT` to the DB.

Some notes about the differences between autocommit mode and our default `REPEATABLE READ` transactions:

1. Currently `autocommit` only applies when using PostgreSQL, and is ignored when using SQLite (due to silliness with [Twisted DB classes](https://twistedmatrix.com/trac/ticket/9998)).
2. Autocommit functions may get retried on error, which means they can get applied *twice* (or more) to the DB (since they are not in a transaction the previous call would not get rolled back). This means that the functions need to be idempotent (or otherwise not care about being called multiple times). Read queries, simple deletes, and updates/upserts that replace rows (rather than generating new values from existing rows) are all idempotent.
3. Autocommit functions no longer get executed in [`REPEATABLE READ`](https://www.postgresql.org/docs/current/transaction-iso.html) isolation level, and so data can change queries, which is fine for single statement queries.

1 files changed, 45 insertions, 31 deletions

diff --git a/synapse/storage/databases/main/transactions.py b/synapse/storage/databases/main/transactions.py
index 7d46090267..59207cadd4 100644
--- a/synapse/storage/databases/main/transactions.py
+++ b/synapse/storage/databases/main/transactions.py
@@ -208,42 +208,56 @@ class TransactionStore(TransactionWorkerStore):
         """
 
         self._destination_retry_cache.pop(destination, None)
-        return await self.db_pool.runInteraction(
-            "set_destination_retry_timings",
-            self._set_destination_retry_timings,
-            destination,
-            failure_ts,
-            retry_last_ts,
-            retry_interval,
-        )
+        if self.database_engine.can_native_upsert:
+            return await self.db_pool.runInteraction(
+                "set_destination_retry_timings",
+                self._set_destination_retry_timings_native,
+                destination,
+                failure_ts,
+                retry_last_ts,
+                retry_interval,
+                db_autocommit=True,  # Safe as its a single upsert
+            )
+        else:
+            return await self.db_pool.runInteraction(
+                "set_destination_retry_timings",
+                self._set_destination_retry_timings_emulated,
+                destination,
+                failure_ts,
+                retry_last_ts,
+                retry_interval,
+            )
 
-    def _set_destination_retry_timings(
+    def _set_destination_retry_timings_native(
         self, txn, destination, failure_ts, retry_last_ts, retry_interval
     ):
+        assert self.database_engine.can_native_upsert
+
+        # Upsert retry time interval if retry_interval is zero (i.e. we're
+        # resetting it) or greater than the existing retry interval.
+        #
+        # WARNING: This is executed in autocommit, so we shouldn't add any more
+        # SQL calls in here (without being very careful).
+        sql = """
+            INSERT INTO destinations (
+                destination, failure_ts, retry_last_ts, retry_interval
+            )
+                VALUES (?, ?, ?, ?)
+            ON CONFLICT (destination) DO UPDATE SET
+                    failure_ts = EXCLUDED.failure_ts,
+                    retry_last_ts = EXCLUDED.retry_last_ts,
+                    retry_interval = EXCLUDED.retry_interval
+                WHERE
+                    EXCLUDED.retry_interval = 0
+                    OR destinations.retry_interval IS NULL
+                    OR destinations.retry_interval < EXCLUDED.retry_interval
+        """
 
-        if self.database_engine.can_native_upsert:
-            # Upsert retry time interval if retry_interval is zero (i.e. we're
-            # resetting it) or greater than the existing retry interval.
-
-            sql = """
-                INSERT INTO destinations (
-                    destination, failure_ts, retry_last_ts, retry_interval
-                )
-                    VALUES (?, ?, ?, ?)
-                ON CONFLICT (destination) DO UPDATE SET
-                        failure_ts = EXCLUDED.failure_ts,
-                        retry_last_ts = EXCLUDED.retry_last_ts,
-                        retry_interval = EXCLUDED.retry_interval
-                    WHERE
-                        EXCLUDED.retry_interval = 0
-                        OR destinations.retry_interval IS NULL
-                        OR destinations.retry_interval < EXCLUDED.retry_interval
-            """
-
-            txn.execute(sql, (destination, failure_ts, retry_last_ts, retry_interval))
-
-            return
+        txn.execute(sql, (destination, failure_ts, retry_last_ts, retry_interval))
 
+    def _set_destination_retry_timings_emulated(
+        self, txn, destination, failure_ts, retry_last_ts, retry_interval
+    ):
         self.database_engine.lock_table(txn, "destinations")
 
         # We need to be careful here as the data may have changed from under us