stats.py 12 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299
  1. # -*- coding: utf-8 -*-
  2. # Copyright 2018 New Vector Ltd
  3. #
  4. # Licensed under the Apache License, Version 2.0 (the "License");
  5. # you may not use this file except in compliance with the License.
  6. # You may obtain a copy of the License at
  7. #
  8. # http://www.apache.org/licenses/LICENSE-2.0
  9. #
  10. # Unless required by applicable law or agreed to in writing, software
  11. # distributed under the License is distributed on an "AS IS" BASIS,
  12. # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  13. # See the License for the specific language governing permissions and
  14. # limitations under the License.
  15. import logging
  16. from collections import Counter
  17. from synapse.api.constants import EventTypes, Membership
  18. from synapse.metrics import event_processing_positions
  19. from synapse.metrics.background_process_metrics import run_as_background_process
  20. logger = logging.getLogger(__name__)
  21. class StatsHandler:
  22. """Handles keeping the *_stats tables updated with a simple time-series of
  23. information about the users, rooms and media on the server, such that admins
  24. have some idea of who is consuming their resources.
  25. Heavily derived from UserDirectoryHandler
  26. """
  27. def __init__(self, hs):
  28. self.hs = hs
  29. self.store = hs.get_datastore()
  30. self.state = hs.get_state_handler()
  31. self.server_name = hs.hostname
  32. self.clock = hs.get_clock()
  33. self.notifier = hs.get_notifier()
  34. self.is_mine_id = hs.is_mine_id
  35. self.stats_bucket_size = hs.config.stats_bucket_size
  36. self.stats_enabled = hs.config.stats_enabled
  37. # The current position in the current_state_delta stream
  38. self.pos = None
  39. # Guard to ensure we only process deltas one at a time
  40. self._is_processing = False
  41. if hs.config.stats_enabled:
  42. self.notifier.add_replication_callback(self.notify_new_event)
  43. # We kick this off so that we don't have to wait for a change before
  44. # we start populating stats
  45. self.clock.call_later(0, self.notify_new_event)
  46. def notify_new_event(self):
  47. """Called when there may be more deltas to process
  48. """
  49. if not self.stats_enabled or self._is_processing:
  50. return
  51. self._is_processing = True
  52. async def process():
  53. try:
  54. await self._unsafe_process()
  55. finally:
  56. self._is_processing = False
  57. run_as_background_process("stats.notify_new_event", process)
  58. async def _unsafe_process(self):
  59. # If self.pos is None then means we haven't fetched it from DB
  60. if self.pos is None:
  61. self.pos = await self.store.get_stats_positions()
  62. # Loop round handling deltas until we're up to date
  63. while True:
  64. # Be sure to read the max stream_ordering *before* checking if there are any outstanding
  65. # deltas, since there is otherwise a chance that we could miss updates which arrive
  66. # after we check the deltas.
  67. room_max_stream_ordering = self.store.get_room_max_stream_ordering()
  68. if self.pos == room_max_stream_ordering:
  69. break
  70. logger.debug(
  71. "Processing room stats %s->%s", self.pos, room_max_stream_ordering
  72. )
  73. max_pos, deltas = await self.store.get_current_state_deltas(
  74. self.pos, room_max_stream_ordering
  75. )
  76. if deltas:
  77. logger.debug("Handling %d state deltas", len(deltas))
  78. room_deltas, user_deltas = await self._handle_deltas(deltas)
  79. else:
  80. room_deltas = {}
  81. user_deltas = {}
  82. # Then count deltas for total_events and total_event_bytes.
  83. (
  84. room_count,
  85. user_count,
  86. ) = await self.store.get_changes_room_total_events_and_bytes(
  87. self.pos, max_pos
  88. )
  89. for room_id, fields in room_count.items():
  90. room_deltas.setdefault(room_id, {}).update(fields)
  91. for user_id, fields in user_count.items():
  92. user_deltas.setdefault(user_id, {}).update(fields)
  93. logger.debug("room_deltas: %s", room_deltas)
  94. logger.debug("user_deltas: %s", user_deltas)
  95. # Always call this so that we update the stats position.
  96. await self.store.bulk_update_stats_delta(
  97. self.clock.time_msec(),
  98. updates={"room": room_deltas, "user": user_deltas},
  99. stream_id=max_pos,
  100. )
  101. logger.debug("Handled room stats to %s -> %s", self.pos, max_pos)
  102. event_processing_positions.labels("stats").set(max_pos)
  103. self.pos = max_pos
  104. async def _handle_deltas(self, deltas):
  105. """Called with the state deltas to process
  106. Returns:
  107. tuple[dict[str, Counter], dict[str, counter]]
  108. Two dicts: the room deltas and the user deltas,
  109. mapping from room/user ID to changes in the various fields.
  110. """
  111. room_to_stats_deltas = {}
  112. user_to_stats_deltas = {}
  113. room_to_state_updates = {}
  114. for delta in deltas:
  115. typ = delta["type"]
  116. state_key = delta["state_key"]
  117. room_id = delta["room_id"]
  118. event_id = delta["event_id"]
  119. stream_id = delta["stream_id"]
  120. prev_event_id = delta["prev_event_id"]
  121. logger.debug("Handling: %r, %r %r, %s", room_id, typ, state_key, event_id)
  122. token = await self.store.get_earliest_token_for_stats("room", room_id)
  123. # If the earliest token to begin from is larger than our current
  124. # stream ID, skip processing this delta.
  125. if token is not None and token >= stream_id:
  126. logger.debug(
  127. "Ignoring: %s as earlier than this room's initial ingestion event",
  128. event_id,
  129. )
  130. continue
  131. if event_id is None and prev_event_id is None:
  132. logger.error(
  133. "event ID is None and so is the previous event ID. stream_id: %s",
  134. stream_id,
  135. )
  136. continue
  137. event_content = {}
  138. sender = None
  139. if event_id is not None:
  140. event = await self.store.get_event(event_id, allow_none=True)
  141. if event:
  142. event_content = event.content or {}
  143. sender = event.sender
  144. # All the values in this dict are deltas (RELATIVE changes)
  145. room_stats_delta = room_to_stats_deltas.setdefault(room_id, Counter())
  146. room_state = room_to_state_updates.setdefault(room_id, {})
  147. if prev_event_id is None:
  148. # this state event doesn't overwrite another,
  149. # so it is a new effective/current state event
  150. room_stats_delta["current_state_events"] += 1
  151. if typ == EventTypes.Member:
  152. # we could use StateDeltasHandler._get_key_change here but it's
  153. # a bit inefficient given we're not testing for a specific
  154. # result; might as well just grab the prev_membership and
  155. # membership strings and compare them.
  156. # We take None rather than leave as a previous membership
  157. # in the absence of a previous event because we do not want to
  158. # reduce the leave count when a new-to-the-room user joins.
  159. prev_membership = None
  160. if prev_event_id is not None:
  161. prev_event = await self.store.get_event(
  162. prev_event_id, allow_none=True
  163. )
  164. if prev_event:
  165. prev_event_content = prev_event.content
  166. prev_membership = prev_event_content.get(
  167. "membership", Membership.LEAVE
  168. )
  169. membership = event_content.get("membership", Membership.LEAVE)
  170. if prev_membership is None:
  171. logger.debug("No previous membership for this user.")
  172. elif membership == prev_membership:
  173. pass # noop
  174. elif prev_membership == Membership.JOIN:
  175. room_stats_delta["joined_members"] -= 1
  176. elif prev_membership == Membership.INVITE:
  177. room_stats_delta["invited_members"] -= 1
  178. elif prev_membership == Membership.LEAVE:
  179. room_stats_delta["left_members"] -= 1
  180. elif prev_membership == Membership.BAN:
  181. room_stats_delta["banned_members"] -= 1
  182. else:
  183. raise ValueError(
  184. "%r is not a valid prev_membership" % (prev_membership,)
  185. )
  186. if membership == prev_membership:
  187. pass # noop
  188. if membership == Membership.JOIN:
  189. room_stats_delta["joined_members"] += 1
  190. elif membership == Membership.INVITE:
  191. room_stats_delta["invited_members"] += 1
  192. if sender and self.is_mine_id(sender):
  193. user_to_stats_deltas.setdefault(sender, Counter())[
  194. "invites_sent"
  195. ] += 1
  196. elif membership == Membership.LEAVE:
  197. room_stats_delta["left_members"] += 1
  198. elif membership == Membership.BAN:
  199. room_stats_delta["banned_members"] += 1
  200. else:
  201. raise ValueError("%r is not a valid membership" % (membership,))
  202. user_id = state_key
  203. if self.is_mine_id(user_id):
  204. # this accounts for transitions like leave → ban and so on.
  205. has_changed_joinedness = (prev_membership == Membership.JOIN) != (
  206. membership == Membership.JOIN
  207. )
  208. if has_changed_joinedness:
  209. delta = +1 if membership == Membership.JOIN else -1
  210. user_to_stats_deltas.setdefault(user_id, Counter())[
  211. "joined_rooms"
  212. ] += delta
  213. room_stats_delta["local_users_in_room"] += delta
  214. elif typ == EventTypes.Create:
  215. room_state["is_federatable"] = (
  216. event_content.get("m.federate", True) is True
  217. )
  218. if sender and self.is_mine_id(sender):
  219. user_to_stats_deltas.setdefault(sender, Counter())[
  220. "rooms_created"
  221. ] += 1
  222. elif typ == EventTypes.JoinRules:
  223. room_state["join_rules"] = event_content.get("join_rule")
  224. elif typ == EventTypes.RoomHistoryVisibility:
  225. room_state["history_visibility"] = event_content.get(
  226. "history_visibility"
  227. )
  228. elif typ == EventTypes.RoomEncryption:
  229. room_state["encryption"] = event_content.get("algorithm")
  230. elif typ == EventTypes.Name:
  231. room_state["name"] = event_content.get("name")
  232. elif typ == EventTypes.Topic:
  233. room_state["topic"] = event_content.get("topic")
  234. elif typ == EventTypes.RoomAvatar:
  235. room_state["avatar"] = event_content.get("url")
  236. elif typ == EventTypes.CanonicalAlias:
  237. room_state["canonical_alias"] = event_content.get("alias")
  238. elif typ == EventTypes.GuestAccess:
  239. room_state["guest_access"] = event_content.get("guest_access")
  240. for room_id, state in room_to_state_updates.items():
  241. logger.debug("Updating room_stats_state for %s: %s", room_id, state)
  242. await self.store.update_room_state(room_id, state)
  243. return room_to_stats_deltas, user_to_stats_deltas