Fix typing being reset causing infinite syncs (#4127)

changelog.d/4127.bugfix

@@ -0,0 +1 @@
+If the typing stream ID goes backwards (as on a worker when the master restarts), the worker's typing handler will no longer erroneously report rooms containing new typing events.

synapse/app/synchrotron.py

@@ -226,7 +226,15 @@ class SynchrotronPresence(object):
 class SynchrotronTyping(object):
     def __init__(self, hs):
         self._latest_room_serial = 0
+        self._reset()
+
+    def _reset(self):
+        """
+        Reset the typing handler's data caches.
+        """
+        # map room IDs to serial numbers
         self._room_serials = {}
+        # map room IDs to sets of users currently typing
         self._room_typing = {}
 
     def stream_positions(self):
@@ -236,6 +244,12 @@ class SynchrotronTyping(object):
         return {"typing": self._latest_room_serial}
 
     def process_replication_rows(self, token, rows):
+        if self._latest_room_serial > token:
+            # The master has gone backwards. To prevent inconsistent data, just
+            # clear everything.
+            self._reset()
+
+        # Set the latest serial token to whatever the server gave us.
         self._latest_room_serial = token
 
         for row in rows:

synapse/handlers/typing.py

@@ -63,11 +63,8 @@ class TypingHandler(object):
         self._member_typing_until = {}  # clock time we expect to stop
         self._member_last_federation_poke = {}
 
-        # map room IDs to serial numbers
-        self._room_serials = {}
         self._latest_room_serial = 0
-        # map room IDs to sets of users currently typing
-        self._room_typing = {}
+        self._reset()
 
         # caches which room_ids changed at which serials
         self._typing_stream_change_cache = StreamChangeCache(
@@ -79,6 +76,15 @@ class TypingHandler(object):
             5000,
         )
 
+    def _reset(self):
+        """
+        Reset the typing handler's data caches.
+        """
+        # map room IDs to serial numbers
+        self._room_serials = {}
+        # map room IDs to sets of users currently typing
+        self._room_typing = {}
+
     def _handle_timeouts(self):
         logger.info("Checking for typing timeouts")
 

tests/rest/client/v2_alpha/test_sync.py

@@ -15,9 +15,11 @@
 
 from mock import Mock
 
+from synapse.rest.client.v1 import admin, login, room
 from synapse.rest.client.v2_alpha import sync
 
 from tests import unittest
+from tests.server import TimedOutException
 
 
 class FilterTestCase(unittest.HomeserverTestCase):
@@ -65,3 +67,124 @@ class FilterTestCase(unittest.HomeserverTestCase):
                 ["next_batch", "rooms", "account_data", "to_device", "device_lists"]
             ).issubset(set(channel.json_body.keys()))
         )
+
+
+class SyncTypingTests(unittest.HomeserverTestCase):
+
+    servlets = [
+        admin.register_servlets,
+        room.register_servlets,
+        login.register_servlets,
+        sync.register_servlets,
+    ]
+    user_id = True
+    hijack_auth = False
+
+    def test_sync_backwards_typing(self):
+        """
+        If the typing serial goes backwards and the typing handler is then reset
+        (such as when the master restarts and sets the typing serial to 0), we
+        do not incorrectly return typing information that had a serial greater
+        than the now-reset serial.
+        """
+        typing_url = "/rooms/%s/typing/%s?access_token=%s"
+        sync_url = "/sync?timeout=3000000&access_token=%s&since=%s"
+
+        # Register the user who gets notified
+        user_id = self.register_user("user", "pass")
+        access_token = self.login("user", "pass")
+
+        # Register the user who sends the message
+        other_user_id = self.register_user("otheruser", "pass")
+        other_access_token = self.login("otheruser", "pass")
+
+        # Create a room
+        room = self.helper.create_room_as(user_id, tok=access_token)
+
+        # Invite the other person
+        self.helper.invite(room=room, src=user_id, tok=access_token, targ=other_user_id)
+
+        # The other user joins
+        self.helper.join(room=room, user=other_user_id, tok=other_access_token)
+
+        # The other user sends some messages
+        self.helper.send(room, body="Hi!", tok=other_access_token)
+        self.helper.send(room, body="There!", tok=other_access_token)
+
+        # Start typing.
+        request, channel = self.make_request(
+            "PUT",
+            typing_url % (room, other_user_id, other_access_token),
+            b'{"typing": true, "timeout": 30000}',
+        )
+        self.render(request)
+        self.assertEquals(200, channel.code)
+
+        request, channel = self.make_request(
+            "GET", "/sync?access_token=%s" % (access_token,)
+        )
+        self.render(request)
+        self.assertEquals(200, channel.code)
+        next_batch = channel.json_body["next_batch"]
+
+        # Stop typing.
+        request, channel = self.make_request(
+            "PUT",
+            typing_url % (room, other_user_id, other_access_token),
+            b'{"typing": false}',
+        )
+        self.render(request)
+        self.assertEquals(200, channel.code)
+
+        # Start typing.
+        request, channel = self.make_request(
+            "PUT",
+            typing_url % (room, other_user_id, other_access_token),
+            b'{"typing": true, "timeout": 30000}',
+        )
+        self.render(request)
+        self.assertEquals(200, channel.code)
+
+        # Should return immediately
+        request, channel = self.make_request(
+            "GET", sync_url % (access_token, next_batch)
+        )
+        self.render(request)
+        self.assertEquals(200, channel.code)
+        next_batch = channel.json_body["next_batch"]
+
+        # Reset typing serial back to 0, as if the master had.
+        typing = self.hs.get_typing_handler()
+        typing._latest_room_serial = 0
+
+        # Since it checks the state token, we need some state to update to
+        # invalidate the stream token.
+        self.helper.send(room, body="There!", tok=other_access_token)
+
+        request, channel = self.make_request(
+            "GET", sync_url % (access_token, next_batch)
+        )
+        self.render(request)
+        self.assertEquals(200, channel.code)
+        next_batch = channel.json_body["next_batch"]
+
+        # This should time out! But it does not, because our stream token is
+        # ahead, and therefore it's saying the typing (that we've actually
+        # already seen) is new, since it's got a token above our new, now-reset
+        # stream token.
+        request, channel = self.make_request(
+            "GET", sync_url % (access_token, next_batch)
+        )
+        self.render(request)
+        self.assertEquals(200, channel.code)
+        next_batch = channel.json_body["next_batch"]
+
+        # Clear the typing information, so that it doesn't think everything is
+        # in the future.
+        typing._reset()
+
+        # Now it SHOULD fail as it never completes!
+        request, channel = self.make_request(
+            "GET", sync_url % (access_token, next_batch)
+        )
+        self.assertRaises(TimedOutException, self.render, request)

tests/server.py

@@ -21,6 +21,12 @@ from synapse.util import Clock
 from tests.utils import setup_test_homeserver as _sth
 
 
+class TimedOutException(Exception):
+    """
+    A web query timed out.
+    """
+
+
 @attr.s
 class FakeChannel(object):
     """
@@ -153,7 +159,7 @@ def wait_until_result(clock, request, timeout=100):
         x += 1
 
         if x > timeout:
-            raise Exception("Timed out waiting for request to finish.")
+            raise TimedOutException("Timed out waiting for request to finish.")
 
         clock.advance(0.1)