Make get_state_groups_from_groups faster.

Most of the time was spent copying a dict to filter out sentinel values
that indicated that keys did not exist in the dict. The sentinel values
were added to ensure that we cached the non-existence of keys.

By updating DictionaryCache to keep track of which keys were known to
not exist itself we can remove a dictionary copy.

1 files changed, 11 insertions, 29 deletions

diff --git a/synapse/storage/state.py b/synapse/storage/state.py

index 85acf2ad1e..a7c3d401d4 100644
--- a/synapse/storage/state.py
+++ b/synapse/storage/state.py
@@ -563,20 +563,22 @@ class StateStore(SQLBaseStore):
                 where a `state_key` of `None` matches all state_keys for the
                 `type`.
         """
-        is_all, state_dict_ids = self._state_group_cache.get(group)
+        is_all, known_absent, state_dict_ids = self._state_group_cache.get(group)
 
         type_to_key = {}
         missing_types = set()
+
         for typ, state_key in types:
+            key = (typ, state_key)
             if state_key is None:
                 type_to_key[typ] = None
-                missing_types.add((typ, state_key))
+                missing_types.add(key)
             else:
                 if type_to_key.get(typ, object()) is not None:
                     type_to_key.setdefault(typ, set()).add(state_key)
 
-                if (typ, state_key) not in state_dict_ids:
-                    missing_types.add((typ, state_key))
+                if key not in state_dict_ids and key not in known_absent:
+                    missing_types.add(key)
 
         sentinel = object()
 
@@ -590,7 +592,7 @@ class StateStore(SQLBaseStore):
                 return True
             return False
 
-        got_all = not (missing_types or types is None)
+        got_all = is_all or not missing_types
 
         return {
             k: v for k, v in state_dict_ids.iteritems()
@@ -607,7 +609,7 @@ class StateStore(SQLBaseStore):
         Args:
             group: The state group to lookup
         """
-        is_all, state_dict_ids = self._state_group_cache.get(group)
+        is_all, _, state_dict_ids = self._state_group_cache.get(group)
 
         return state_dict_ids, is_all
 
@@ -624,7 +626,7 @@ class StateStore(SQLBaseStore):
         missing_groups = []
         if types is not None:
             for group in set(groups):
-                state_dict_ids, missing_types, got_all = self._get_some_state_from_cache(
+                state_dict_ids, _, got_all = self._get_some_state_from_cache(
                     group, types
                 )
                 results[group] = state_dict_ids
@@ -653,19 +655,7 @@ class StateStore(SQLBaseStore):
             # Now we want to update the cache with all the things we fetched
             # from the database.
             for group, group_state_dict in group_to_state_dict.iteritems():
-                if types:
-                    # We delibrately put key -> None mappings into the cache to
-                    # cache absence of the key, on the assumption that if we've
-                    # explicitly asked for some types then we will probably ask
-                    # for them again.
-                    state_dict = {
-                        (intern_string(etype), intern_string(state_key)): None
-                        for (etype, state_key) in types
-                    }
-                    state_dict.update(results[group])
-                    results[group] = state_dict
-                else:
-                    state_dict = results[group]
+                state_dict = results[group]
 
                 state_dict.update(
                     ((intern_string(k[0]), intern_string(k[1])), to_ascii(v))
@@ -677,17 +667,9 @@ class StateStore(SQLBaseStore):
                     key=group,
                     value=state_dict,
                     full=(types is None),
+                    known_absent=types,
                 )
 
-        # Remove all the entries with None values. The None values were just
-        # used for bookkeeping in the cache.
-        for group, state_dict in results.iteritems():
-            results[group] = {
-                key: event_id
-                for key, event_id in state_dict.iteritems()
-                if event_id
-            }
-
         defer.returnValue(results)
 
     def get_next_state_group(self):