Improve exception handling for background processes

There were a bunch of places where we fire off a process to happen in the
background, but don't have any exception handling on it - instead relying on
the unhandled error being logged when the relevent deferred gets
garbage-collected.

This is unsatisfactory for a number of reasons:
 - logging on garbage collection is best-effort and may happen some time after
   the error, if at all
 - it can be hard to figure out where the error actually happened.
 - it is logged as a scary CRITICAL error which (a) I always forget to grep for
   and (b) it's not really CRITICAL if a background process we don't care about
   fails.

So this is an attempt to add exception handling to everything we fire off into
the background.

1 files changed, 49 insertions, 46 deletions

diff --git a/synapse/app/synchrotron.py b/synapse/app/synchrotron.py
index 2fddcd935a..777da564d7 100644
--- a/synapse/app/synchrotron.py
+++ b/synapse/app/synchrotron.py
@@ -340,55 +340,58 @@ class SyncReplicationHandler(ReplicationClientHandler):
 
     @defer.inlineCallbacks
     def process_and_notify(self, stream_name, token, rows):
-        if stream_name == "events":
-            # We shouldn't get multiple rows per token for events stream, so
-            # we don't need to optimise this for multiple rows.
-            for row in rows:
-                event = yield self.store.get_event(row.event_id)
-                extra_users = ()
-                if event.type == EventTypes.Member:
-                    extra_users = (event.state_key,)
-                max_token = self.store.get_room_max_stream_ordering()
-                self.notifier.on_new_room_event(
-                    event, token, max_token, extra_users
+        try:
+            if stream_name == "events":
+                # We shouldn't get multiple rows per token for events stream, so
+                # we don't need to optimise this for multiple rows.
+                for row in rows:
+                    event = yield self.store.get_event(row.event_id)
+                    extra_users = ()
+                    if event.type == EventTypes.Member:
+                        extra_users = (event.state_key,)
+                    max_token = self.store.get_room_max_stream_ordering()
+                    self.notifier.on_new_room_event(
+                        event, token, max_token, extra_users
+                    )
+            elif stream_name == "push_rules":
+                self.notifier.on_new_event(
+                    "push_rules_key", token, users=[row.user_id for row in rows],
                 )
-        elif stream_name == "push_rules":
-            self.notifier.on_new_event(
-                "push_rules_key", token, users=[row.user_id for row in rows],
-            )
-        elif stream_name in ("account_data", "tag_account_data",):
-            self.notifier.on_new_event(
-                "account_data_key", token, users=[row.user_id for row in rows],
-            )
-        elif stream_name == "receipts":
-            self.notifier.on_new_event(
-                "receipt_key", token, rooms=[row.room_id for row in rows],
-            )
-        elif stream_name == "typing":
-            self.typing_handler.process_replication_rows(token, rows)
-            self.notifier.on_new_event(
-                "typing_key", token, rooms=[row.room_id for row in rows],
-            )
-        elif stream_name == "to_device":
-            entities = [row.entity for row in rows if row.entity.startswith("@")]
-            if entities:
+            elif stream_name in ("account_data", "tag_account_data",):
                 self.notifier.on_new_event(
-                    "to_device_key", token, users=entities,
+                    "account_data_key", token, users=[row.user_id for row in rows],
                 )
-        elif stream_name == "device_lists":
-            all_room_ids = set()
-            for row in rows:
-                room_ids = yield self.store.get_rooms_for_user(row.user_id)
-                all_room_ids.update(room_ids)
-            self.notifier.on_new_event(
-                "device_list_key", token, rooms=all_room_ids,
-            )
-        elif stream_name == "presence":
-            yield self.presence_handler.process_replication_rows(token, rows)
-        elif stream_name == "receipts":
-            self.notifier.on_new_event(
-                "groups_key", token, users=[row.user_id for row in rows],
-            )
+            elif stream_name == "receipts":
+                self.notifier.on_new_event(
+                    "receipt_key", token, rooms=[row.room_id for row in rows],
+                )
+            elif stream_name == "typing":
+                self.typing_handler.process_replication_rows(token, rows)
+                self.notifier.on_new_event(
+                    "typing_key", token, rooms=[row.room_id for row in rows],
+                )
+            elif stream_name == "to_device":
+                entities = [row.entity for row in rows if row.entity.startswith("@")]
+                if entities:
+                    self.notifier.on_new_event(
+                        "to_device_key", token, users=entities,
+                    )
+            elif stream_name == "device_lists":
+                all_room_ids = set()
+                for row in rows:
+                    room_ids = yield self.store.get_rooms_for_user(row.user_id)
+                    all_room_ids.update(room_ids)
+                self.notifier.on_new_event(
+                    "device_list_key", token, rooms=all_room_ids,
+                )
+            elif stream_name == "presence":
+                yield self.presence_handler.process_replication_rows(token, rows)
+            elif stream_name == "receipts":
+                self.notifier.on_new_event(
+                    "groups_key", token, users=[row.user_id for row in rows],
+                )
+        except Exception:
+            logger.exception("Error processing replication")
 
 
 def start(config_options):