Don't go into federation catch up mode so easily

Federation catch up mode is very inefficient if the number of events
that the remote server has missed is small, since handling gaps can be
very expensive, c.f. #9492.

Instead of going into catch up mode whenever we see an error, we instead
do so only if we've backed off from trying the remote for more than an
hour (the assumption being that in such a case it is more than a
transient failure).

synapse/federation/sender/per_destination_queue.py

@@ -401,7 +401,7 @@ async def _transaction_transmission_loop(self) -> None:
                 self._pending_presence = {}
                 self._pending_rrs = {}
 
-            self._start_catching_up()
+                self._start_catching_up()
         except FederationDeniedError as e:
             logger.info(e)
         except HttpResponseException as e:
@@ -412,7 +412,6 @@ async def _transaction_transmission_loop(self) -> None:
                 e,
             )
 
-            self._start_catching_up()
         except RequestSendFailed as e:
             logger.warning(
                 "TX [%s] Failed to send transaction: %s", self._destination, e
@@ -422,16 +421,12 @@ async def _transaction_transmission_loop(self) -> None:
                 logger.info(
                     "Failed to send event %s to %s", p.event_id, self._destination
                 )
-
-            self._start_catching_up()
         except Exception:
             logger.exception("TX [%s] Failed to send transaction", self._destination)
             for p in pending_pdus:
                 logger.info(
                     "Failed to send event %s to %s", p.event_id, self._destination
                 )
-
-            self._start_catching_up()
         finally:
             # We want to be *very* sure we clear this after we stop processing
             self.transmission_loop_running = False