diff options
Diffstat (limited to 'synapse/storage/event_push_actions.py')
| -rw-r--r-- | synapse/storage/event_push_actions.py | 406 |
1 files changed, 293 insertions, 113 deletions
diff --git a/synapse/storage/event_push_actions.py b/synapse/storage/event_push_actions.py index 575d710d5d..e78f8d0114 100644 --- a/synapse/storage/event_push_actions.py +++ b/synapse/storage/event_push_actions.py @@ -1,5 +1,6 @@ # -*- coding: utf-8 -*- # Copyright 2015 OpenMarket Ltd +# Copyright 2018 New Vector Ltd # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. @@ -13,7 +14,7 @@ # See the License for the specific language governing permissions and # limitations under the License. -from ._base import SQLBaseStore +from synapse.storage._base import SQLBaseStore, LoggingTransaction from twisted.internet import defer from synapse.util.async import sleep from synapse.util.caches.descriptors import cachedInlineCallbacks @@ -62,60 +63,28 @@ def _deserialize_action(actions, is_highlight): return DEFAULT_NOTIF_ACTION -class EventPushActionsStore(SQLBaseStore): - EPA_HIGHLIGHT_INDEX = "epa_highlight_index" - +class EventPushActionsWorkerStore(SQLBaseStore): def __init__(self, db_conn, hs): - super(EventPushActionsStore, self).__init__(db_conn, hs) - - self.register_background_index_update( - self.EPA_HIGHLIGHT_INDEX, - index_name="event_push_actions_u_highlight", - table="event_push_actions", - columns=["user_id", "stream_ordering"], + super(EventPushActionsWorkerStore, self).__init__(db_conn, hs) + + # These get correctly set by _find_stream_orderings_for_times_txn + self.stream_ordering_month_ago = None + self.stream_ordering_day_ago = None + + cur = LoggingTransaction( + db_conn.cursor(), + name="_find_stream_orderings_for_times_txn", + database_engine=self.database_engine, + after_callbacks=[], + exception_callbacks=[], ) + self._find_stream_orderings_for_times_txn(cur) + cur.close() - self.register_background_index_update( - "event_push_actions_highlights_index", - index_name="event_push_actions_highlights_index", - table="event_push_actions", - columns=["user_id", "room_id", "topological_ordering", "stream_ordering"], - where_clause="highlight=1" + self.find_stream_orderings_looping_call = self._clock.looping_call( + self._find_stream_orderings_for_times, 10 * 60 * 1000 ) - self._doing_notif_rotation = False - self._rotate_notif_loop = self._clock.looping_call( - self._rotate_notifs, 30 * 60 * 1000 - ) - - def _set_push_actions_for_event_and_users_txn(self, txn, event, tuples): - """ - Args: - event: the event set actions for - tuples: list of tuples of (user_id, actions) - """ - values = [] - for uid, actions in tuples: - is_highlight = 1 if _action_has_highlight(actions) else 0 - - values.append({ - 'room_id': event.room_id, - 'event_id': event.event_id, - 'user_id': uid, - 'actions': _serialize_action(actions, is_highlight), - 'stream_ordering': event.internal_metadata.stream_ordering, - 'topological_ordering': event.depth, - 'notif': 1, - 'highlight': is_highlight, - }) - - for uid, __ in tuples: - txn.call_after( - self.get_unread_event_push_actions_by_room_for_user.invalidate_many, - (event.room_id, uid) - ) - self._simple_insert_many_txn(txn, "event_push_actions", values) - @cachedInlineCallbacks(num_args=3, tree=True, max_entries=5000) def get_unread_event_push_actions_by_room_for_user( self, room_id, user_id, last_read_event_id @@ -432,6 +401,280 @@ class EventPushActionsStore(SQLBaseStore): # Now return the first `limit` defer.returnValue(notifs[:limit]) + def add_push_actions_to_staging(self, event_id, user_id_actions): + """Add the push actions for the event to the push action staging area. + + Args: + event_id (str) + user_id_actions (dict[str, list[dict|str])]): A dictionary mapping + user_id to list of push actions, where an action can either be + a string or dict. + + Returns: + Deferred + """ + + if not user_id_actions: + return + + # This is a helper function for generating the necessary tuple that + # can be used to inert into the `event_push_actions_staging` table. + def _gen_entry(user_id, actions): + is_highlight = 1 if _action_has_highlight(actions) else 0 + return ( + event_id, # event_id column + user_id, # user_id column + _serialize_action(actions, is_highlight), # actions column + 1, # notif column + is_highlight, # highlight column + ) + + def _add_push_actions_to_staging_txn(txn): + # We don't use _simple_insert_many here to avoid the overhead + # of generating lists of dicts. + + sql = """ + INSERT INTO event_push_actions_staging + (event_id, user_id, actions, notif, highlight) + VALUES (?, ?, ?, ?, ?) + """ + + txn.executemany(sql, ( + _gen_entry(user_id, actions) + for user_id, actions in user_id_actions.iteritems() + )) + + return self.runInteraction( + "add_push_actions_to_staging", _add_push_actions_to_staging_txn + ) + + def remove_push_actions_from_staging(self, event_id): + """Called if we failed to persist the event to ensure that stale push + actions don't build up in the DB + + Args: + event_id (str) + """ + + return self._simple_delete( + table="event_push_actions_staging", + keyvalues={ + "event_id": event_id, + }, + desc="remove_push_actions_from_staging", + ) + + @defer.inlineCallbacks + def _find_stream_orderings_for_times(self): + yield self.runInteraction( + "_find_stream_orderings_for_times", + self._find_stream_orderings_for_times_txn + ) + + def _find_stream_orderings_for_times_txn(self, txn): + logger.info("Searching for stream ordering 1 month ago") + self.stream_ordering_month_ago = self._find_first_stream_ordering_after_ts_txn( + txn, self._clock.time_msec() - 30 * 24 * 60 * 60 * 1000 + ) + logger.info( + "Found stream ordering 1 month ago: it's %d", + self.stream_ordering_month_ago + ) + logger.info("Searching for stream ordering 1 day ago") + self.stream_ordering_day_ago = self._find_first_stream_ordering_after_ts_txn( + txn, self._clock.time_msec() - 24 * 60 * 60 * 1000 + ) + logger.info( + "Found stream ordering 1 day ago: it's %d", + self.stream_ordering_day_ago + ) + + def find_first_stream_ordering_after_ts(self, ts): + """Gets the stream ordering corresponding to a given timestamp. + + Specifically, finds the stream_ordering of the first event that was + received on or after the timestamp. This is done by a binary search on + the events table, since there is no index on received_ts, so is + relatively slow. + + Args: + ts (int): timestamp in millis + + Returns: + Deferred[int]: stream ordering of the first event received on/after + the timestamp + """ + return self.runInteraction( + "_find_first_stream_ordering_after_ts_txn", + self._find_first_stream_ordering_after_ts_txn, + ts, + ) + + @staticmethod + def _find_first_stream_ordering_after_ts_txn(txn, ts): + """ + Find the stream_ordering of the first event that was received on or + after a given timestamp. This is relatively slow as there is no index + on received_ts but we can then use this to delete push actions before + this. + + received_ts must necessarily be in the same order as stream_ordering + and stream_ordering is indexed, so we manually binary search using + stream_ordering + + Args: + txn (twisted.enterprise.adbapi.Transaction): + ts (int): timestamp to search for + + Returns: + int: stream ordering + """ + txn.execute("SELECT MAX(stream_ordering) FROM events") + max_stream_ordering = txn.fetchone()[0] + + if max_stream_ordering is None: + return 0 + + # We want the first stream_ordering in which received_ts is greater + # than or equal to ts. Call this point X. + # + # We maintain the invariants: + # + # range_start <= X <= range_end + # + range_start = 0 + range_end = max_stream_ordering + 1 + + # Given a stream_ordering, look up the timestamp at that + # stream_ordering. + # + # The array may be sparse (we may be missing some stream_orderings). + # We treat the gaps as the same as having the same value as the + # preceding entry, because we will pick the lowest stream_ordering + # which satisfies our requirement of received_ts >= ts. + # + # For example, if our array of events indexed by stream_ordering is + # [10, <none>, 20], we should treat this as being equivalent to + # [10, 10, 20]. + # + sql = ( + "SELECT received_ts FROM events" + " WHERE stream_ordering <= ?" + " ORDER BY stream_ordering DESC" + " LIMIT 1" + ) + + while range_end - range_start > 0: + middle = (range_end + range_start) // 2 + txn.execute(sql, (middle,)) + row = txn.fetchone() + if row is None: + # no rows with stream_ordering<=middle + range_start = middle + 1 + continue + + middle_ts = row[0] + if ts > middle_ts: + # we got a timestamp lower than the one we were looking for. + # definitely need to look higher: X > middle. + range_start = middle + 1 + else: + # we got a timestamp higher than (or the same as) the one we + # were looking for. We aren't yet sure about the point we + # looked up, but we can be sure that X <= middle. + range_end = middle + + return range_end + + +class EventPushActionsStore(EventPushActionsWorkerStore): + EPA_HIGHLIGHT_INDEX = "epa_highlight_index" + + def __init__(self, db_conn, hs): + super(EventPushActionsStore, self).__init__(db_conn, hs) + + self.register_background_index_update( + self.EPA_HIGHLIGHT_INDEX, + index_name="event_push_actions_u_highlight", + table="event_push_actions", + columns=["user_id", "stream_ordering"], + ) + + self.register_background_index_update( + "event_push_actions_highlights_index", + index_name="event_push_actions_highlights_index", + table="event_push_actions", + columns=["user_id", "room_id", "topological_ordering", "stream_ordering"], + where_clause="highlight=1" + ) + + self._doing_notif_rotation = False + self._rotate_notif_loop = self._clock.looping_call( + self._rotate_notifs, 30 * 60 * 1000 + ) + + def _set_push_actions_for_event_and_users_txn(self, txn, events_and_contexts, + all_events_and_contexts): + """Handles moving push actions from staging table to main + event_push_actions table for all events in `events_and_contexts`. + + Also ensures that all events in `all_events_and_contexts` are removed + from the push action staging area. + + Args: + events_and_contexts (list[(EventBase, EventContext)]): events + we are persisting + all_events_and_contexts (list[(EventBase, EventContext)]): all + events that we were going to persist. This includes events + we've already persisted, etc, that wouldn't appear in + events_and_context. + """ + + sql = """ + INSERT INTO event_push_actions ( + room_id, event_id, user_id, actions, stream_ordering, + topological_ordering, notif, highlight + ) + SELECT ?, event_id, user_id, actions, ?, ?, notif, highlight + FROM event_push_actions_staging + WHERE event_id = ? + """ + + if events_and_contexts: + txn.executemany(sql, ( + ( + event.room_id, event.internal_metadata.stream_ordering, + event.depth, event.event_id, + ) + for event, _ in events_and_contexts + )) + + for event, _ in events_and_contexts: + user_ids = self._simple_select_onecol_txn( + txn, + table="event_push_actions_staging", + keyvalues={ + "event_id": event.event_id, + }, + retcol="user_id", + ) + + for uid in user_ids: + txn.call_after( + self.get_unread_event_push_actions_by_room_for_user.invalidate_many, + (event.room_id, uid,) + ) + + # Now we delete the staging area for *all* events that were being + # persisted. + txn.executemany( + "DELETE FROM event_push_actions_staging WHERE event_id = ?", + ( + (event.event_id,) + for event, _ in all_events_and_contexts + ) + ) + @defer.inlineCallbacks def get_push_actions_for_user(self, user_id, before=None, limit=50, only_highlight=False): @@ -551,69 +794,6 @@ class EventPushActionsStore(SQLBaseStore): """, (room_id, user_id, stream_ordering)) @defer.inlineCallbacks - def _find_stream_orderings_for_times(self): - yield self.runInteraction( - "_find_stream_orderings_for_times", - self._find_stream_orderings_for_times_txn - ) - - def _find_stream_orderings_for_times_txn(self, txn): - logger.info("Searching for stream ordering 1 month ago") - self.stream_ordering_month_ago = self._find_first_stream_ordering_after_ts_txn( - txn, self._clock.time_msec() - 30 * 24 * 60 * 60 * 1000 - ) - logger.info( - "Found stream ordering 1 month ago: it's %d", - self.stream_ordering_month_ago - ) - logger.info("Searching for stream ordering 1 day ago") - self.stream_ordering_day_ago = self._find_first_stream_ordering_after_ts_txn( - txn, self._clock.time_msec() - 24 * 60 * 60 * 1000 - ) - logger.info( - "Found stream ordering 1 day ago: it's %d", - self.stream_ordering_day_ago - ) - - def _find_first_stream_ordering_after_ts_txn(self, txn, ts): - """ - Find the stream_ordering of the first event that was received after - a given timestamp. This is relatively slow as there is no index on - received_ts but we can then use this to delete push actions before - this. - - received_ts must necessarily be in the same order as stream_ordering - and stream_ordering is indexed, so we manually binary search using - stream_ordering - """ - txn.execute("SELECT MAX(stream_ordering) FROM events") - max_stream_ordering = txn.fetchone()[0] - - if max_stream_ordering is None: - return 0 - - range_start = 0 - range_end = max_stream_ordering - - sql = ( - "SELECT received_ts FROM events" - " WHERE stream_ordering > ?" - " ORDER BY stream_ordering" - " LIMIT 1" - ) - - while range_end - range_start > 1: - middle = int((range_end + range_start) / 2) - txn.execute(sql, (middle,)) - middle_ts = txn.fetchone()[0] - if ts > middle_ts: - range_start = middle - else: - range_end = middle - - return range_end - - @defer.inlineCallbacks def _rotate_notifs(self): if self._doing_notif_rotation or self.stream_ordering_day_ago is None: return |