Reduce size of joined_user cache

The _get_joined_users_from_context cache stores a mapping from user_id
to avatar_url and display_name. Instead of storing those in a dict,
store them in a namedtuple as that uses much less memory.

We also try converting the string to ascii to further reduce the size.

4 files changed, 40 insertions, 11 deletions

diff --git a/synapse/push/bulk_push_rule_evaluator.py b/synapse/push/bulk_push_rule_evaluator.py
index 78b095c903..503fef4261 100644
--- a/synapse/push/bulk_push_rule_evaluator.py
+++ b/synapse/push/bulk_push_rule_evaluator.py
@@ -87,8 +87,11 @@ class BulkPushRuleEvaluator:
         condition_cache = {}
 
         for uid, rules in self.rules_by_user.items():
-            display_name = room_members.get(uid, {}).get("display_name", None)
-            if not display_name:
+            display_name = None
+            profile_info = room_members.get(uid, {})
+            if profile_info:
+                display_name = profile_info.display_name
+            else:
                 # Handle the case where we are pushing a membership event to
                 # that user, as they might not be already joined.
                 if event.type == EventTypes.Member and event.state_key == uid:
diff --git a/synapse/rest/client/v1/room.py b/synapse/rest/client/v1/room.py
index 0bdd6b5b36..c376ab8fd7 100644
--- a/synapse/rest/client/v1/room.py
+++ b/synapse/rest/client/v1/room.py
@@ -406,7 +406,13 @@ class JoinedRoomMemberListRestServlet(ClientV1RestServlet):
         users_with_profile = yield self.state.get_current_user_in_room(room_id)
 
         defer.returnValue((200, {
-            "joined": users_with_profile
+            "joined": {
+                user_id: {
+                    "avatar_url": profile.avatar_url,
+                    "display_name": profile.display_name,
+                }
+                for user_id, profile in users_with_profile.iteritems()
+            }
         }))
 
 
diff --git a/synapse/storage/roommember.py b/synapse/storage/roommember.py
index 367dbbbcf6..68e724ac7f 100644
--- a/synapse/storage/roommember.py
+++ b/synapse/storage/roommember.py
@@ -19,6 +19,7 @@ from collections import namedtuple
 
 from ._base import SQLBaseStore
 from synapse.util.caches.descriptors import cached, cachedInlineCallbacks
+from synapse.util.stringutils import to_ascii
 
 from synapse.api.constants import Membership, EventTypes
 from synapse.types import get_domain_from_id
@@ -35,6 +36,11 @@ RoomsForUser = namedtuple(
 )
 
 
+ProfileInfo = namedtuple(
+    "ProfileInfo", ("avatar_url", "display_name")
+)
+
+
 _MEMBERSHIP_PROFILE_UPDATE_NAME = "room_membership_profile_update"
 
 
@@ -422,20 +428,20 @@ class RoomMemberStore(SQLBaseStore):
         )
 
         users_in_room = {
-            row["user_id"]: {
-                "display_name": row["display_name"],
-                "avatar_url": row["avatar_url"],
-            }
+            to_ascii(row["user_id"]): ProfileInfo(
+                avatar_url=to_ascii(row["avatar_url"]),
+                display_name=to_ascii(row["display_name"]),
+            )
             for row in rows
         }
 
         if event is not None and event.type == EventTypes.Member:
             if event.membership == Membership.JOIN:
                 if event.event_id in member_event_ids:
-                    users_in_room[event.state_key] = {
-                        "display_name": event.content.get("displayname", None),
-                        "avatar_url": event.content.get("avatar_url", None),
-                    }
+                    users_in_room[to_ascii(event.state_key)] = ProfileInfo(
+                        display_name=to_ascii(event.content.get("displayname", None)),
+                        avatar_url=to_ascii(event.content.get("avatar_url", None)),
+                    )
 
         defer.returnValue(users_in_room)
 
diff --git a/synapse/util/stringutils.py b/synapse/util/stringutils.py
index a100f151d4..95a6168e16 100644
--- a/synapse/util/stringutils.py
+++ b/synapse/util/stringutils.py
@@ -40,3 +40,17 @@ def is_ascii(s):
         return False
     else:
         return True
+
+
+def to_ascii(s):
+    """Converts a string to ascii if it is ascii, otherwise leave it alone.
+
+    If given None then will return None.
+    """
+    if s is None:
+        return None
+
+    try:
+        return s.encode("ascii")
+    except UnicodeEncodeError:
+        return s