Then adapts calling code to retry when needed so it doesn't 500
to clients.
Signed-off-by: Mathieu Velten <mathieuv@matrix.org>
Co-authored-by: Sean Quah <8349537+squahtx@users.noreply.github.com>
We don't filter state usually, so doing so here is a waste of time. This is not much of an issue for clients that enable lazy loading of members, since there will be fewer state events.
* Introduce a test for the old behaviour which we want to restore
* Reintroduce the old behaviour in a simpler way
* Newsfile
Signed-off-by: Olivier Wilkinson (reivilibre) <oliverw@matrix.org>
* Use 1 credit instead of 2 for creating a room: be more lenient than before
Notably, the UI in Element Web was still broken after restoring to prior behaviour.
After discussion, we agreed that it would be sensible to increase the limit.
Signed-off-by: Olivier Wilkinson (reivilibre) <oliverw@matrix.org>
1. `federation_client.timestamp_to_event(...)` now handles all `destination` looping and uses our generic `_try_destination_list(...)` helper.
2. Consistently handling `NotRetryingDestination` and `FederationDeniedError` across `get_pdu` , backfill, and the generic `_try_destination_list` which is used for many places we use this pattern.
3. `get_pdu(...)` now returns `PulledPduInfo` so we know which `destination` we ended up pulling the PDU from
The callers either set a default limit or manually handle a None-limit
later on (by setting a default value).
Update the callers to always instantiate PaginationConfig with a default
limit and then assume the limit is non-None.
Fixes#13942. Introduced in #13575.
Basically, let's only get the ordered set of hosts out of the DB if we need an ordered set of hosts. Since we split the function up the caching won't be as good, but I think it will still be fine as e.g. multiple backfill requests for the same room will hit the cache.
This PR begins work on batching up events during the creation of a room. The PR splits out the creation and sending/persisting of the events. The first three events in the creation of the room-creating the room, joining the creator to the room, and the power levels event are sent sequentially, while the subsequent events are created and collected to be sent at the end of the function. This is currently done by appending them to a list and then iterating over the list to send, the next step (after this PR) would be to send and persist the collected events as a batch.
Optimize how we calculate `likely_domains` during backfill because I've seen this take 17s in production just to `get_current_state` which is used to `get_domains_from_state` (see case [*2. Loading tons of events* in the `/messages` investigation issue](https://github.com/matrix-org/synapse/issues/13356)).
There are 3 ways we currently calculate hosts that are in the room:
1. `get_current_state` -> `get_domains_from_state`
- Used in `backfill` to calculate `likely_domains` and `/timestamp_to_event` because it was cargo-culted from `backfill`
- This one is being eliminated in favor of `get_current_hosts_in_room` in this PR 🕳
1. `get_current_hosts_in_room`
- Used for other federation things like sending read receipts and typing indicators
1. `get_hosts_in_room_at_events`
- Used when pushing out events over federation to other servers in the `_process_event_queue_loop`
Fix https://github.com/matrix-org/synapse/issues/13626
Part of https://github.com/matrix-org/synapse/issues/13356
Mentioned in [internal doc](https://docs.google.com/document/d/1lvUoVfYUiy6UaHB6Rb4HicjaJAU40-APue9Q4vzuW3c/edit#bookmark=id.2tvwz3yhcafh)
### Query performance
#### Before
The query from `get_current_state` sucks just because we have to get all 80k events. And we see almost the exact same performance locally trying to get all of these events (16s vs 17s):
```
synapse=# SELECT type, state_key, event_id FROM current_state_events WHERE room_id = '!OGEhHVWSdvArJzumhm:matrix.org';
Time: 16035.612 ms (00:16.036)
synapse=# SELECT type, state_key, event_id FROM current_state_events WHERE room_id = '!OGEhHVWSdvArJzumhm:matrix.org';
Time: 4243.237 ms (00:04.243)
```
But what about `get_current_hosts_in_room`: When there is 8M rows in the `current_state_events` table, the previous query in `get_current_hosts_in_room` took 13s from complete freshness (when the events were first added). But takes 930ms after a Postgres restart or 390ms if running back to back to back.
```sh
$ psql synapse
synapse=# \timing on
synapse=# SELECT COUNT(DISTINCT substring(state_key FROM '@[^:]*:(.*)$'))
FROM current_state_events
WHERE
type = 'm.room.member'
AND membership = 'join'
AND room_id = '!OGEhHVWSdvArJzumhm:matrix.org';
count
-------
4130
(1 row)
Time: 13181.598 ms (00:13.182)
synapse=# SELECT COUNT(*) from current_state_events where room_id = '!OGEhHVWSdvArJzumhm:matrix.org';
count
-------
80814
synapse=# SELECT COUNT(*) from current_state_events;
count
---------
8162847
synapse=# SELECT pg_size_pretty( pg_total_relation_size('current_state_events') );
pg_size_pretty
----------------
4702 MB
```
#### After
I'm not sure how long it takes from complete freshness as I only really get that opportunity once (maybe restarting computer but that's cumbersome) and it's not really relevant to normal operating times. Maybe you get closer to the fresh times the more access variability there is so that Postgres caches aren't as exact. Update: The longest I've seen this run for is 6.4s and 4.5s after a computer restart.
After a Postgres restart, it takes 330ms and running back to back takes 260ms.
```sh
$ psql synapse
synapse=# \timing on
Timing is on.
synapse=# SELECT
substring(c.state_key FROM '@[^:]*:(.*)$') as host
FROM current_state_events c
/* Get the depth of the event from the events table */
INNER JOIN events AS e USING (event_id)
WHERE
c.type = 'm.room.member'
AND c.membership = 'join'
AND c.room_id = '!OGEhHVWSdvArJzumhm:matrix.org'
GROUP BY host
ORDER BY min(e.depth) ASC;
Time: 333.800 ms
```
#### Going further
To improve things further we could add a `limit` parameter to `get_current_hosts_in_room`. Realistically, we don't need 4k domains to choose from because there is no way we're going to query that many before we a) probably get an answer or b) we give up.
Another thing we can do is optimize the query to use a index skip scan:
- https://wiki.postgresql.org/wiki/Loose_indexscan
- Index Skip Scan, https://commitfest.postgresql.org/37/1741/
- https://www.timescale.com/blog/how-we-made-distinct-queries-up-to-8000x-faster-on-postgresql/
Part of #13019
This changes all the permission-related methods to rely on the Requester instead of the UserID. This is a first step towards enabling scoped access tokens at some point, since I expect the Requester to have scope-related informations in it.
It also changes methods which figure out the user/device/appservice out of the access token to return a Requester instead of something else. This avoids having store-related objects in the methods signatures.
Inspired by the room batch handler, this uses previous event inserts to
pre-populate prev events during room creation, reducing the number of
queries required to create a room.
Signed off by Nick @ Beeper (@Fizzadar)
Complement tests: https://github.com/matrix-org/complement/pull/405
This happens when you have some messages imported before the room is created.
Then use MSC3030 to look backwards before the room creation from a remote
federated server. The server won't find anything locally, but will ask over
federation which will have the remote event. The previous logic would
choke on not having the local event assigned.
```
Failed to fetch /timestamp_to_event from hs2 because of exception(UnboundLocalError) local variable 'local_event' referenced before assignment args=("local variable 'local_event' referenced before assignment",)
```
Instead, use the `room_version` property of the event we're validating.
The `room_version` was originally added as a parameter somewhere around #4482,
but really it's been redundant since #6875 added a `room_version` field to `EventBase`.
* Update worker docs to remove group endpoints.
* Removes an unused parameter to `ApplicationService`.
* Break dependency between media repo and groups.
* Avoid copying `m.room.related_groups` state events during room upgrades.
Refactor and convert `Linearizer` to async. This makes a `Linearizer`
cancellation bug easier to fix.
Also refactor to use an async context manager, which eliminates an
unlikely footgun where code that doesn't immediately use the context
manager could forget to release the lock.
Signed-off-by: Sean Quah <seanq@element.io>
This is some odds and ends found during the review of #11791
and while continuing to work in this code:
* Return attrs classes instead of dictionaries from some methods
to improve type safety.
* Call `get_bundled_aggregations` fewer times.
* Adds a missing assertion in the tests.
* Do not return empty bundled aggregations for an event (preferring
to not include the bundle at all, as the docstring states).
This makes the serialization of events synchronous (and it no
longer access the database), but we must manually calculate and
provide the bundled aggregations.
Overall this should cause no change in behavior, but is prep work
for other improvements.
Mathieu Velten