Cache literal sync filter validation (#17186)

The sliding sync proxy (amongst other things) use literal json blobs as
filters, and repeatedly validating them takes a bunch of CPU.
This commit is contained in:
Erik Johnston 2024-05-14 15:08:46 +01:00 committed by GitHub
parent ebe77381b0
commit 284d85dee3
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
2 changed files with 14 additions and 1 deletions

1
changelog.d/17186.misc Normal file
View file

@ -0,0 +1 @@
Cache literal sync filter validation for performance.

View file

@ -47,6 +47,7 @@ from synapse.http.site import SynapseRequest
from synapse.logging.opentracing import trace_with_opname
from synapse.types import JsonDict, Requester, StreamToken
from synapse.util import json_decoder
from synapse.util.caches.lrucache import LruCache
from ._base import client_patterns, set_timeline_upper_limit
@ -110,6 +111,11 @@ class SyncRestServlet(RestServlet):
self._msc2654_enabled = hs.config.experimental.msc2654_enabled
self._msc3773_enabled = hs.config.experimental.msc3773_enabled
self._json_filter_cache: LruCache[str, bool] = LruCache(
max_size=1000,
cache_name="sync_valid_filter",
)
async def on_GET(self, request: SynapseRequest) -> Tuple[int, JsonDict]:
# This will always be set by the time Twisted calls us.
assert request.args is not None
@ -177,7 +183,13 @@ class SyncRestServlet(RestServlet):
filter_object = json_decoder.decode(filter_id)
except Exception:
raise SynapseError(400, "Invalid filter JSON", errcode=Codes.NOT_JSON)
self.filtering.check_valid_filter(filter_object)
# We cache the validation, as this can get quite expensive if people use
# a literal json blob as a query param.
if not self._json_filter_cache.get(filter_id):
self.filtering.check_valid_filter(filter_object)
self._json_filter_cache[filter_id] = True
set_timeline_upper_limit(
filter_object, self.hs.config.server.filter_timeline_limit
)