Fix have_seen_event cache not being invalidated (#13863)

Fix https://github.com/matrix-org/synapse/issues/13856
Fix https://github.com/matrix-org/synapse/issues/13865

> Discovered while trying to make Synapse fast enough for [this MSC2716 test for importing many batches](https://github.com/matrix-org/complement/pull/214#discussion_r741678240). As an example, disabling the `have_seen_event` cache saves 10 seconds for each `/messages` request in that MSC2716 Complement test because we're not making as many federation requests for `/state` (speeding up `have_seen_event` itself is related to https://github.com/matrix-org/synapse/issues/13625) 
> 
> But this will also make `/messages` faster in general so we can include it in the [faster `/messages` milestone](https://github.com/matrix-org/synapse/milestone/11).
> 
> *-- https://github.com/matrix-org/synapse/issues/13856*


### The problem

`_invalidate_caches_for_event` doesn't run in monolith mode which means we never even tried to clear the `have_seen_event` and other caches. And even in worker mode, it only runs on the workers, not the master (AFAICT).

Additionally there was bug with the key being wrong so `_invalidate_caches_for_event` never invalidates the `have_seen_event` cache even when it does run.

Because we were using the `@cachedList` wrong, it was putting items in the cache under keys like `((room_id, event_id),)` with a `set` in a `set` (ex. `(('!TnCIJPKzdQdUlIyXdQ:test', '$Iu0eqEBN7qcyF1S9B3oNB3I91v2o5YOgRNPwi_78s-k'),)`) and we we're trying to invalidate with just `(room_id, event_id)` which did nothing.
This commit is contained in:
Eric Eastwood 2022-09-27 15:55:43 -05:00 committed by GitHub
parent 35e9d6a616
commit 29269d9d3f
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
5 changed files with 165 additions and 67 deletions

View file

@ -13,7 +13,7 @@
# See the License for the specific language governing permissions and
# limitations under the License.
import logging
from typing import Set
from typing import Iterable, Set, Tuple
from unittest import mock
from twisted.internet import defer, reactor
@ -1008,3 +1008,34 @@ class CachedListDescriptorTestCase(unittest.TestCase):
obj.inner_context_was_finished, "Tried to restart a finished logcontext"
)
self.assertEqual(current_context(), SENTINEL_CONTEXT)
def test_num_args_mismatch(self):
"""
Make sure someone does not accidentally use @cachedList on a method with
a mismatch in the number args to the underlying single cache method.
"""
class Cls:
@descriptors.cached(tree=True)
def fn(self, room_id, event_id):
pass
# This is wrong ❌. `@cachedList` expects to be given the same number
# of arguments as the underlying cached function, just with one of
# the arguments being an iterable
@descriptors.cachedList(cached_method_name="fn", list_name="keys")
def list_fn(self, keys: Iterable[Tuple[str, str]]):
pass
# Corrected syntax ✅
#
# @cachedList(cached_method_name="fn", list_name="event_ids")
# async def list_fn(
# self, room_id: str, event_ids: Collection[str],
# )
obj = Cls()
# Make sure this raises an error about the arg mismatch
with self.assertRaises(Exception):
obj.list_fn([("foo", "bar")])