summary refs log tree commit diff
path: root/synapse (follow)
Commit message (Collapse)AuthorAgeFilesLines
* Bump flake8-bugbear from 22.9.23 to 22.10.27 (#14329)dependabot[bot]2022-11-043-4/+6
| | | | | Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: GitHub Actions <github-actions[bot]@users.noreply.github.com> Co-authored-by: Olivier Wilkinson (reivilibre) <oliverw@matrix.org>
* Fix /refresh endpoint version (#14364)Tulir Asokan2022-11-041-1/+1
|
* Merge tag 'v1.71.0rc2' into developSean Quah2022-11-041-1/+1
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Synapse 1.71.0rc2 (2022-11-04) ============================== Please note that, as announced in the release notes for Synapse 1.69.0, legacy Prometheus metric names are now disabled by default. They will be removed altogether in Synapse 1.73.0. If not already done, server administrators should update their dashboards and alerting rules to avoid using the deprecated metric names. See the [upgrade notes](https://matrix-org.github.io/synapse/v1.71/upgrade.html#upgrading-to-v1710) for more details. Improved Documentation ---------------------- - Document the changes to monthly active user metrics due to deprecation of legacy Prometheus metric names. ([\#14358](https://github.com/matrix-org/synapse/issues/14358), [\#14360](https://github.com/matrix-org/synapse/issues/14360)) Deprecations and Removals ------------------------- - Disable legacy Prometheus metric names by default. They can still be re-enabled for now, but they will be removed altogether in Synapse 1.73.0. ([\#14353](https://github.com/matrix-org/synapse/issues/14353)) Internal Changes ---------------- - Run unit tests against Python 3.11. ([\#13812](https://github.com/matrix-org/synapse/issues/13812))
| * Disable legacy Prometheus metric names by default. They can still be ↵reivilibre2022-11-021-1/+1
| | | | | | | | re-enabled for now, but they will be removed altogether in Synapse 1.73.0. (#14353)
* | Implement MSC3912: Relation-based redactions (#14260)Brendan Abolivier2022-11-037-24/+179
| | | | | | Co-authored-by: Sean Quah <8349537+squahtx@users.noreply.github.com>
* | Fix typo in #13320 which could cause log spam (#14347)David Robertson2022-11-011-1/+1
| |
* | Allow PUT/GET of aliases during faster join (#14292)David Robertson2022-11-011-2/+2
|/ | | without blocking on full state.
* Revert "Fix event size checks (#13710)"David Robertson2022-11-011-5/+5
| | | | | | | | | | | This reverts commit fab495a9e1442d99e922367f65f41de5eaa488eb. As noted in https://github.com/matrix-org/synapse/pull/13710#issuecomment-1298396007: > We want to see this change land for the protocol's sake (and plan to un-revert it) but want to give this a little more time before releasing this.
* Fix type annotation causing import time error in the Complement forking ↵reivilibre2022-11-011-2/+2
| | | | | launcher. (#14084) Co-authored-by: David Robertson <davidr@element.io>
* Fix dehydrated device REST checks (#14336)David Robertson2022-10-311-3/+2
|
* Support OIDC backchannel logouts (#11414)Quentin Gliech2022-10-316-36/+488
| | | | | | | If configured an OIDC IdP can log a user's session out of Synapse when they log out of the identity provider. The IdP sends a request directly to Synapse (and must be configured with an endpoint) when a user logs out.
* Avoid incrementing bg process utime/stime counters by negative durations ↵David Robertson2022-10-311-2/+4
| | | | (#14323)
* Prevent federation user keys query from returning device names if disallowed ↵Andrew Morgan2022-10-282-9/+45
| | | | (#14304)
* Improve `RawHeaders` type hints (#14303)David Robertson2022-10-282-9/+23
|
* Switch search SQL to triple-quote strings. (#14311)Patrick Cloke2022-10-281-89/+99
| | | | For ease of reading we switch from concatenated strings to triple quote strings.
* Merge branch 'master' into developOlivier Wilkinson (reivilibre)2022-10-283-11/+25
|\
| * Fix room creation being rate limited too aggressively since Synapse v1.69.0. ↵reivilibre2022-10-282-6/+18
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | (#14314) * 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>
| * Fix incorrectly sending authentication tokens to application service as ↵David Robertson2022-10-261-5/+7
| | | | | | | | headers (#14301)
* | Check appservice user interest against the local users instead of all users ↵Eric Eastwood2022-10-273-4/+32
| | | | | | | | (`get_users_in_room` mis-use) (#13958)
* | Fix tests for change in PostgreSQL 14 behavior change. (#14310)Patrick Cloke2022-10-271-3/+2
| | | | | | | | | | | | | | PostgreSQL 14 changed the behavior of `websearch_to_tsquery` to improve some behaviour. The tests were hitting those edge-cases about handling of hanging double quotes. This fixes the tests to take into account the PostgreSQL version.
* | Add workers settings to configuration manual (#14086)Dirk Klimpel2022-10-271-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * Add workers settings to configuration manual * Update `pusher_instances` * update url to python logger * update headlines * update links after headline change * remove link from `daemon process` There is no docs in Synapse for this * extend example for `federation_sender_instances` and `pusher_instances` * more infos about stream writers * add link to DAG * update `pusher_instances` * update `worker_listeners` * update `stream_writers` * Update `worker_name` Co-authored-by: David Robertson <davidr@element.io>
* | Fix presence bug introduced in 1.64 by #13313 (#14243)Mathieu Velten2022-10-271-2/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | * Fix presence bug introduced in 1.64 by #13313 Signed-off-by: Mathieu Velten <mathieuv@matrix.org> * Add changelog * Add DISTINCT * Apply suggestions from code review Signed-off-by: Mathieu Velten <mathieuv@matrix.org>
* | Reject history insertion during partial joins (#14291)David Robertson2022-10-271-0/+7
| |
* | Refactor MSC3030 `/timestamp_to_event` to move away from our snowflake pull ↵Eric Eastwood2022-10-265-122/+182
| | | | | | | | | | | | | | | | | | from `destination` pattern (#14096) 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
* | fix broken avatar checks when server_name contains a port (#13927)Ashish Kumar2022-10-261-1/+5
| | | | | | | | | | | | Fixes check_avatar_size_and_mime_type() to successfully update avatars on homeservers running on non-default ports which it would mistakenly treat as remote homeserver while validating the avatar's size and mime type. Signed-off-by: Ashish Kumar ashfame@users.noreply.github.com
* | Save login tokens in database (#13844)Quentin Gliech2022-10-267-142/+249
| | | | | | | | | | | | | | | | | | | | | | * Save login tokens in database Signed-off-by: Quentin Gliech <quenting@element.io> * Add upgrade notes * Track login token reuse in a Prometheus metric Signed-off-by: Quentin Gliech <quenting@element.io>
* | Unified search query syntax using the full-text search capabilities of the ↵James Salter2022-10-253-35/+240
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | underlying DB. (#11635) Support a unified search query syntax which leverages more of the full-text search of each database supported by Synapse. Supports, with the same syntax across Postgresql 11+ and Sqlite: - quoted "search terms" - `AND`, `OR`, `-` (negation) operators - Matching words based on their stem, e.g. searches for "dog" matches documents containing "dogs". This is achieved by - If on postgresql 11+, pass the user input to `websearch_to_tsquery` - If on sqlite, manually parse the query and transform it into the sqlite-specific query syntax. Note that postgresql 10, which is close to end-of-life, falls back to using `phraseto_tsquery`, which only supports a subset of the features. Multiple terms separated by a space are implicitly ANDed. Note that: 1. There is no escaping of full-text syntax that might be supported by the database; e.g. `NOT`, `NEAR`, `*` in sqlite. This runs the risk that people might discover this as accidental functionality and depend on something we don't guarantee. 2. English text is assumed for stemming. To support other languages, either the target language needs to be known at the time of indexing the message (via room metadata, or otherwise), or a separate index for each language supported could be created. Sqlite docs: https://www.sqlite.org/fts3.html#full_text_index_queries Postgres docs: https://www.postgresql.org/docs/11/textsearch-controls.html
* | Merge branch 'release-v1.70' into developOlivier Wilkinson (reivilibre)2022-10-253-16/+59
|\|
| * Revert behavior change for bundling edits of non-message events (#14283)Patrick Cloke2022-10-241-4/+7
| |
| * Fix `TypeError: 'dict_keys' object is not reversible` (#14280)Erik Johnston2022-10-241-2/+1
| |
| * Properly update the threads table when thread events are redacted. (#14248)Patrick Cloke2022-10-211-10/+51
| | | | | | | | | | | | | | | | | | When the last event in a thread is redacted we need to update the threads table: * Find the new latest event in the thread and store it into the table; or * Remove the thread from the table if it is no longer a thread (i.e. all events in the thread were redacted).
* | Refactor OIDC tests to better mimic an actual OIDC provider. (#13910)Quentin Gliech2022-10-251-4/+11
| | | | | | | | | | | | | | | | | | This implements a fake OIDC server, which intercepts calls to the HTTP client. Improves accuracy of tests by covering more internal methods. One particular example was the ID token validation, which previously mocked. This uncovered an incorrect dependency: Synapse actually requires at least authlib 0.15.1, not 0.14.0.
* | Implementation for MSC3664: Pushrules for relations (#11804)DeepBlueV7.X2022-10-254-5/+67
| |
* | Remove unused `@lru_cache` decorator (#13595)Nick Mills-Barrett2022-10-251-104/+0
| | | | | | | | | | | | | | * Remove unused `@lru_cache` decorator Spotted this working on something else. Co-authored-by: David Robertson <davidr@element.io>
* | Enable WAL for SQLite (#13897)asymmetric2022-10-251-0/+4
| | | | | | Signed-off-by: Lorenzo Manacorda <lorenzo@mailbox.org>
* | Return NOT_JSON if decode fails and defer set_timeline_upper_limit ca… ↵Ryan Miguel2022-10-241-4/+4
| | | | | | | | | | | | | | (#14262) * Return NOT_JSON if decode fails and defer set_timeline_upper_limit call until after check_valid_filter. Fixes #13661. Signed-off-by: Ryan Miguel <miguel.ryanj@gmail.com>. * Reword changelog
* | Add debugging to help diagnose lost device-list-update (#14268)Richard van der Hoff2022-10-241-17/+37
| |
* | Add initial power level event to batch of bulk persisted events when ↵Shay2022-10-215-56/+79
| | | | | | | | creating a new room. (#14228)
* | Improve aesthetics and reusability of HTML templates. (#13652)Germain2022-10-2130-740/+690
| | | | | | | | | | | | | | Use a base template to create a cohesive feel across the HTML templates provided by Synapse. Adds basic styling to the base template for a more user-friendly look and feel.
* | Fix logging error on SIGHUP (#14258)Richard van der Hoff2022-10-211-1/+1
| |
* | Show erasure status when listing users in the Admin API (#14205)Tadeusz Sośnierz2022-10-212-2/+12
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * Show erasure status when listing users in the Admin API * Use USING when joining erased_users * Add changelog entry * Revert "Use USING when joining erased_users" This reverts commit 30bd2bf106415caadcfdbdd1b234ef2b106cc394. * Make the erased check work on postgres * Add a testcase for showing erased user status * Appease the style linter * Explicitly convert `erased` to bool to make SQLite consistent with Postgres This also adds us an easy way in to fix the other accidentally integered columns. * Move erasure status test to UsersListTestCase * Include user erased status when fetching user info via the admin API * Document the erase status in user_admin_api * Appease the linter and mypy * Signpost comments in tests Co-authored-by: Tadeusz Sośnierz <tadeusz@sosnierz.com> Co-authored-by: David Robertson <david.m.robertson1@gmail.com>
* | Fix event size checks (#13710)DeepBlueV7.X2022-10-211-5/+5
| |
* | Use servlets for /key/ endpoints. (#14229)Patrick Cloke2022-10-206-80/+82
| | | | | | | | | | To fix the response for unknown endpoints under that prefix. See MSC3743.
* | Stop returning `unsigned.invite_room_state` in `PUT ↵Andrew Morgan2022-10-201-0/+5
| | | | | | | | | | /_matrix/federation/v2/invite/{roomId}/{eventId}` responses (#14064) Co-authored-by: David Robertson <davidr@element.io>
* | Explain `SynapseError` and `FederationError` better (#14191)Eric Eastwood2022-10-192-3/+29
| | | | | | | | | | Explain `SynapseError` and `FederationError` better Spawning from https://github.com/matrix-org/synapse/pull/13816#discussion_r993262622
* | Bump flake8-bugbear from 21.3.2 to 22.9.23 (#14042)dependabot[bot]2022-10-193-5/+5
| | | | | | | | | | Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: Erik Johnston <erik@matrix.org> Co-authored-by: David Robertson <davidr@element.io>
* | let update_synapse_database run on a multi-database configurations (#13422)Finn2022-10-191-8/+0
|/ | | | | | | | | | | | | | | | | * Allow sharded database in db migrate script Signed-off-by: Finn Herzfeld <finn@beeper.com> * Update changelog.d/13422.bugfix Co-authored-by: Patrick Cloke <clokep@users.noreply.github.com> * Remove check entirely * remove unused import Signed-off-by: Finn Herzfeld <finn@beeper.com> Co-authored-by: finn <finn@beeper.com> Co-authored-by: Patrick Cloke <clokep@users.noreply.github.com>
* Fix MSC3030 `/timestamp_to_event` returning `outliers` that it has no idea ↵Eric Eastwood2022-10-181-21/+38
| | | | | | | | | | | | | | | | | | | | | | | | | | | whether are near a gap or not (#14215) Fix MSC3030 `/timestamp_to_event` endpoint returning `outliers` that it has no idea whether are near a gap or not (and therefore unable to determine whether it's actually the closest event). The reason Synapse doesn't know whether an `outlier` is next to a gap is because our gap checks rely on entries in the `event_edges`, `event_forward_extremeties`, and `event_backward_extremities` tables which is [not the case for `outliers`](https://github.com/matrix-org/synapse/blob/2c63cdcc3f1aa4625e947de3c23e0a8133c61286/docs/development/room-dag-concepts.md#outliers). Also fixes MSC3030 Complement `can_paginate_after_getting_remote_event_from_timestamp_to_event_endpoint` test flake. Although this acted flakey in Complement, if `sync_partial_state` raced and beat us before `/timestamp_to_event`, then even if we retried the failing `/context` request it wouldn't work until we made this Synapse change. With this PR, Synapse will never return an `outlier` event so that test will always go and ask over federation. Fix https://github.com/matrix-org/synapse/issues/13944 ### Why did this fail before? Why was it flakey? Sleuthing the server logs on the [CI failure](https://github.com/matrix-org/synapse/actions/runs/3149623842/jobs/5121449357#step:5:5805), it looks like `hs2:/timestamp_to_event` found `$NP6-oU7mIFVyhtKfGvfrEQX949hQX-T-gvuauG6eurU` as an `outlier` event locally. Then when we went and asked for it via `/context`, since it's an `outlier`, it was filtered out of the results -> `You don't have permission to access that event.` This is reproducible when `sync_partial_state` races and persists `$NP6-oU7mIFVyhtKfGvfrEQX949hQX-T-gvuauG6eurU` as an `outlier` before we evaluate `get_event_for_timestamp(...)`. To consistently reproduce locally, just add a delay at the [start of `get_event_for_timestamp(...)`](https://github.com/matrix-org/synapse/blob/cb20b885cb4bd1648581dd043a184d86fc8c7a00/synapse/handlers/room.py#L1470-L1496) so it always runs after `sync_partial_state` completes. ```py from twisted.internet import task as twisted_task d = twisted_task.deferLater(self.hs.get_reactor(), 3.5) await d ``` In a run where it passes, on `hs2`, `get_event_for_timestamp(...)` finds a different event locally which is next to a gap and we request from a closer one from `hs1` which gets backfilled. And since the backfilled event is not an `outlier`, it's returned as expected during `/context`. With this PR, Synapse will never return an `outlier` event so that test will always go and ask over federation.
* Fix hiding devices names over federation (#10015)Aaron Raimist2022-10-181-2/+8
| | | | | | And don't include blank opentracing stuff in device list updates. Signed-off-by: Aaron Raimist <aaron@raim.ist>
* Fix docstring in EventContext (#14145)Shay2022-10-181-1/+2
|
* Prepatory work for adding power level event to batched events (#14214)Shay2022-10-185-17/+46
|
* Implementation of HTTP 307 response for MSC3886 POST endpoint (#14018)Hugh Nimmo-Smith2022-10-1811-17/+136
| | | | Co-authored-by: reivilibre <olivier@librepush.net> Co-authored-by: Andrew Morgan <andrewm@element.io>
* Update the thread_id right before use (in case the bg update hasn't ↵Patrick Cloke2022-10-184-134/+111
| | | | | | | | finished) (#14222) This avoids running a forced-update of a null thread_id rows. An index is added (in the background) to hopefully make this easier in the future.
* Fix `track_memory_usage` on poetry-core 1.3.x installations (#14221)David Robertson2022-10-181-1/+1
| | | | | | | | | | | | | * Fix `track_memory_usage` on poetry-core 1.3.x installations The same kind of problem as discussed in #14085: 1. we defined an extra with an underscore 2. we look it up at runtime with an underscore 3. but poetry-core 1.3.x. installs it with a dash, causing (2) to fail. Fix by using a dash everywhere. * Changelog
* When restarting a partial join resync, prioritise the server which actioned ↵David Robertson2022-10-185-31/+94
| | | | a partial join (#14126)
* Avoid checking the event cache when backfilling events (#14164)Andrew Morgan2022-10-182-14/+35
|
* Remove `_get_events_cache` check optimisation from `_have_seen_events_dict` ↵Andrew Morgan2022-10-181-18/+13
| | | | (#14161)
* Add debug logs to figure out why an event was filtered (#14095)Eric Eastwood2022-10-171-1/+31
| | | | | | | | | | | | | | | | Spawned while investigating https://github.com/matrix-org/synapse/issues/13944 This way we might get some more context whenever an `403 Forbidden - body: {"errcode":"M_FORBIDDEN","error":"You don't have permission to access that event."}` error is produced. `log_config.yaml` ```yaml loggers: synapse: level: INFO synapse.visibility: level: DEBUG ```
* Support filtering the /messages API by relation type (MSC3874). (#14148)Patrick Cloke2022-10-174-5/+56
| | | Gated behind an experimental configuration flag.
* Use Pydantic when PUTting room aliases (#14179)David Robertson2022-10-172-31/+46
|
* Invalidate rooms for user caches when receiving membership events (#14155)Nick Mills-Barrett2022-10-171-0/+4
| | | | | This should fix a race where the event notification comes in over replication before the state replication, leaving a window during which a sync may get an incorrect list of rooms for the user.
* Stop getting missing `prev_events` after we already know their signature is ↵Eric Eastwood2022-10-154-0/+122
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | invalid (#13816) While https://github.com/matrix-org/synapse/pull/13635 stops us from doing the slow thing after we've already done it once, this PR stops us from doing one of the slow things in the first place. Related to - https://github.com/matrix-org/synapse/issues/13622 - https://github.com/matrix-org/synapse/pull/13635 - https://github.com/matrix-org/synapse/issues/13676 Part of https://github.com/matrix-org/synapse/issues/13356 Follow-up to https://github.com/matrix-org/synapse/pull/13815 which tracks event signature failures. With this PR, we avoid the call to the costly `_get_state_ids_after_missing_prev_event` because the signature failure will count as an attempt before and we filter events based on the backoff before calling `_get_state_ids_after_missing_prev_event` now. For example, this will save us 156s out of the 185s total that this `matrix.org` `/messages` request. If you want to see the full Jaeger trace of this, you can drag and drop this `trace.json` into your own Jaeger, https://gist.github.com/MadLittleMods/4b12d0d0afe88c2f65ffcc907306b761 To explain this exact scenario around `/messages` -> backfill, we call `/backfill` and first check the signatures of the 100 events. We see bad signature for `$luA4l7QHhf_jadH3mI-AyFqho0U2Q-IXXUbGSMq6h6M` and `$zuOn2Rd2vsC7SUia3Hp3r6JSkSFKcc5j3QTTqW_0jDw` (both member events). Then we process the 98 events remaining that have valid signatures but one of the events references `$luA4l7QHhf_jadH3mI-AyFqho0U2Q-IXXUbGSMq6h6M` as a `prev_event`. So we have to do the whole `_get_state_ids_after_missing_prev_event` rigmarole which pulls in those same events which fail again because the signatures are still invalid. - `backfill` - `outgoing-federation-request` `/backfill` - `_check_sigs_and_hash_and_fetch` - `_check_sigs_and_hash_and_fetch_one` for each event received over backfill - ❗ `$luA4l7QHhf_jadH3mI-AyFqho0U2Q-IXXUbGSMq6h6M` fails with `Signature on retrieved event was invalid.`: `unable to verify signature for sender domain xxx: 401: Failed to find any key to satisfy: _FetchKeyRequest(...)` - ❗ `$zuOn2Rd2vsC7SUia3Hp3r6JSkSFKcc5j3QTTqW_0jDw` fails with `Signature on retrieved event was invalid.`: `unable to verify signature for sender domain xxx: 401: Failed to find any key to satisfy: _FetchKeyRequest(...)` - `_process_pulled_events` - `_process_pulled_event` for each validated event - ❗ Event `$Q0iMdqtz3IJYfZQU2Xk2WjB5NDF8Gg8cFSYYyKQgKJ0` references `$luA4l7QHhf_jadH3mI-AyFqho0U2Q-IXXUbGSMq6h6M` as a `prev_event` which is missing so we try to get it - `_get_state_ids_after_missing_prev_event` - `outgoing-federation-request` `/state_ids` - ❗ `get_pdu` for `$luA4l7QHhf_jadH3mI-AyFqho0U2Q-IXXUbGSMq6h6M` which fails the signature check again - ❗ `get_pdu` for `$zuOn2Rd2vsC7SUia3Hp3r6JSkSFKcc5j3QTTqW_0jDw` which fails the signature check
* Merge remote-tracking branch 'origin/release-v1.69' into developPatrick Cloke2022-10-143-15/+77
|\
| * Fix background update to use an index (#14181)Erik Johnston2022-10-141-11/+51
| |
| * Optimise the event_push_backfill_thread_id bg job (#14172)David Robertson2022-10-131-2/+8
| | | | | | Co-authored-by: Erik Johnston <erik@matrix.org>
| * Fix rotating existing notifications in push summary (#14138)Erik Johnston2022-10-111-5/+15
| | | | | | | | | | Broke by #14045. Fixes #14120. Introduced in v1.69.0rc2.
| * Fallback if 'approved' isn't included in a registration replication request ↵Brendan Abolivier2022-10-111-1/+17
| | | | | | | | (#14135)
| * Unpin build-system requirements, but impose an upper-bound (#14085)David Robertson2022-10-071-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * Revert to prior build-system requirements This reverts #14080. * Use normalised extra name, which poetry-core 1.3 will generate anyway * Changelog * Upper bound build-system requirements * Remove upgrade note; expand changelog entry a little. * Fix typo in build-system comment Co-authored-by: Richard van der Hoff <1389908+richvdh@users.noreply.github.com> Co-authored-by: Richard van der Hoff <1389908+richvdh@users.noreply.github.com>
* | Accept threaded receipts for events related to the root event. (#14174)Patrick Cloke2022-10-143-8/+135
| | | | | | | | | | | | | | | | | | The root node of a thread (and events related to it) are considered "part of a thread" when validating receipts. This allows clients which show the root node in both the main timeline and the threaded timeline to easily send receipts in either. Note that threaded notifications are not created for these events, these events created notifications on the main timeline.
* | Advertise support for Matrix 1.4. (#14184)Patrick Cloke2022-10-141-0/+1
| | | | | | | | All features / changes in Matrix 1.4 are now supported in Synapse.
* | Don't require optional `invite_room_state` field on fed v2 invite (#14083)Andrew Morgan2022-10-141-1/+1
| |
* | Do not allow a None-limit on PaginationConfig. (#14146)Patrick Cloke2022-10-1414-49/+26
| | | | | | | | | | | | | | 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.
* | Stabilize the threads API. (#14175)Patrick Cloke2022-10-142-10/+2
| | | | | | | | | | | | | | Stabilize the threads API (MSC3856) by supporting (only) the v1 path for the endpoint. This also marks the API as safe for workers since it is a read-only API.
* | Properly invalidate get_thread_id cache. (#14163)Patrick Cloke2022-10-141-0/+1
| | | | | | This was missed in 2b6d41ebd685fb546e52acdbcb0024dfcf5a5db1 (#13824).
* | Fix sqlite syntax for upserts. (#14171)Patrick Cloke2022-10-131-1/+1
| |
* | Properly return the thread ID down sync. (#14159)Patrick Cloke2022-10-131-2/+2
| | | | | | | | | | Fix a broken conflict in e6e876b9b158f47811b6dfedd8783f658ce960a4, by not stomping over a field right after creating it.
* | Add an API for listing threads in a room. (#13394)Patrick Cloke2022-10-138-6/+370
| | | | | | | | | | | | | | | | | | Implement the /threads endpoint from MSC3856. This is currently unstable and behind an experimental configuration flag. It includes a background update to backfill data, results from the /threads endpoint will be partial until that finishes.
* | Fix a bug where the joined hosts for a given event were not being properly ↵Shay2022-10-122-45/+50
| | | | | | | | cached (#14125)
* | Return the thread ID properly down sync. (#14159)Patrick Cloke2022-10-121-0/+2
| | | | | | | | | | A receipt's thread ID, if one exists, should be added to the body of a receipt.
* | Return the main timeline for events which are not part of a thread. (#14140)Patrick Cloke2022-10-122-6/+8
| | | | | | | | Fixes a bug where threaded receipts could not be sent for the main timeline.
* | Correct field name for stripped state events when knocking. ↵Andrew Morgan2022-10-123-6/+25
| | | | | | | | `knock_state_events` -> `knock_room_state` (#14102)
* | Mark /relations endpoint as usable on workers. (#14028)Patrick Cloke2022-10-121-0/+2
| | | | | | Co-authored-by: Eric Eastwood <erice@element.io>
* | Batch up calls to `get_rooms_for_users` (#14109)Nick Mills-Barrett2022-10-121-1/+16
| |
* | Remove the experimental implementation of MSC3772. (#14094)Patrick Cloke2022-10-126-135/+7
| | | | | | MSC3772 has been abandoned.
* | Fix a bug where redactions were not being sent over federation if we did not ↵Shay2022-10-114-31/+50
| | | | | | | | have the original event. (#13813)
* | Remove the groups config code. (#14142)Patrick Cloke2022-10-111-27/+0
| | | | | | This has been unused for a long time, but missed removal in #11584.
* | Making parse_server_name more consistent (#14007)Abdullah Osama2022-10-111-2/+2
| | | | | | Fixes #12122
* | Indicate what endpoint came back with a JSON response we were unable to ↵Eric Eastwood2022-10-071-1/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | parse (#14097) **Before:** ``` WARNING - POST-11 - Unable to parse JSON: Expecting value: line 1 column 1 (char 0) (b'') ``` **After:** ``` WARNING - POST-11 - Unable to parse JSON from POST /_matrix/client/v3/join/%21ZlmJtelqFroDRJYZaq:hs1?server_name=hs1 response: Expecting value: line 1 column 1 (char 0) (b'') ``` --- It's possible to figure out which endpoint these warnings were coming from before but you had to follow the request ID `POST-11` to the log line that says `Completed request [...]`. Including this key information next to the JSON parsing error makes it much easier to reason whether it matters or not. ``` 2022-09-29T08:23:25.7875506Z synapse_main | 2022-09-29 08:21:10,336 - synapse.http.matrixfederationclient - 299 - INFO - POST-11 - {GET-O-13} [hs1] Completed request: 200 OK in 0.53 secs, got 450 bytes - GET matrix://hs1/_matrix/federation/v1/make_join/%21ohtKoQiXlPePSycXwp%3Ahs1/%40charlie%3Ahs2?ver=1&ver=2&ver=3&ver=4&ver=5&ver=6&ver=org.matrix.msc2176&ver=7&ver=8&ver=9&ver=org.matrix.msc3787&ver=10&ver=org.matrix.msc2716v4 ``` --- As a note, having no `body` is normal for the `/join` endpoint and it can handle it. https://github.com/matrix-org/synapse/blob/0c853e09709d52783efd37060ed9e8f55a4fc704/synapse/rest/client/room.py#L398-L403 Alternatively we could remove these extra logs but they are probably more usually helpful to figure out what went wrong.
* | Remove support for the unstable dir flag on relations. (#14106)Patrick Cloke2022-10-074-57/+30
| | | | | | | | | | | | From MSC3715, this was unused by clients (and there was no way for clients to know it was supported). Matrix 1.4 defines the stable field.
* | Apply & bundle edits for non-message events. (#14034)Patrick Cloke2022-10-071-7/+4
| | | | | | | | | | | | | | Fixes two related bugs: * No edit information was bundled for events which aren't `m.room.message`. * `m.new_content` was not applied for those events.
* | Parse SYNAPSE_ASYNC_IO_REACTOR env variable & log the reactor on startup ↵Quentin Gliech2022-10-072-14/+15
| | | | | | | | (#14092)
* | Be more lenient in the oEmbed response parsing. (#14089)Patrick Cloke2022-10-071-50/+57
| | | | | | | | | | | | Attempt to parse any valid information from an oEmbed response (instead of bailing at the first unexpected data). This should allow for more partial oEmbed data to be returned, resulting in better / more URL previews, even if those URL previews are only partial.
* | Use stable identifiers for MSC3771 & MSC3773. (#14050)Patrick Cloke2022-10-077-48/+44
| | | | | | | | | | These are both part of Matrix 1.4 which has now been released. For now, support both the unstable and stable identifiers.
* | Use Pydantic to validate /devices endpoints (#14054)David Robertson2022-10-071-46/+52
| |
* | Catch BrokenPipeError from metrics server, and log as a warning (#14072)David Robertson2022-10-071-6/+12
| |
* | Always close _all_ `ijson` coroutines, even if doing so raises Exceptions ↵David Robertson2022-10-062-5/+38
| | | | | | | | (#14065)
* | Merge tag 'v1.69.0rc2' into developSean Quah2022-10-062-1/+50
|\| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Synapse 1.69.0rc2 (2022-10-06) ============================== Please note that legacy Prometheus metric names are now deprecated and will be removed in Synapse 1.73.0. Server administrators should update their dashboards and alerting rules to avoid using the deprecated metric names. See the [upgrade notes](https://matrix-org.github.io/synapse/v1.69/upgrade.html#upgrading-to-v1690) for more details. Deprecations and Removals ------------------------- - Deprecate the `generate_short_term_login_token` method in favor of an async `create_login_token` method in the Module API. ([\#13842](https://github.com/matrix-org/synapse/issues/13842)) Internal Changes ---------------- - Ensure Synapse v1.69 works with upcoming database changes in v1.70. ([\#14045](https://github.com/matrix-org/synapse/issues/14045)) - Fix a bug introduced in Synapse v1.68.0 where messages could not be sent in rooms with non-integer `notifications` power level. ([\#14073](https://github.com/matrix-org/synapse/issues/14073)) - Temporarily pin build-system requirements to workaround an incompatibility with poetry-core 1.3.0. This will be reverted before the v1.69.0 release proper, see [\#14079](https://github.com/matrix-org/synapse/issues/14079). ([\#14080](https://github.com/matrix-org/synapse/issues/14080))
| * Fix sending events into rooms with non-integer power levels (#14073)David Robertson2022-10-061-1/+8
| |
| * Deprecate the `generate_short_term_login_token` method in favor of an async ↵Quentin Gliech2022-10-061-0/+42
| | | | | | | | | | | | `create_login_token` method in the Module API. (#13842) Signed-off-by: Quentin Gliech <quenting@element.io> Co-authored-by: Brendan Abolivier <babolivier@matrix.org>
| * Fix backwards compatibility with upcoming threads schema changes. (#14045)Patrick Cloke2022-10-051-11/+23
| | | | | | | | Ensure that the upsert will work properly by first updating any existing rows (in the same way that the background update to backfill data works).
* | Batch up notifications after event persistence (#14033)Shay2022-10-054-58/+65
| |
* | Fix handling of public rooms filter with a network tuple. (#14053)Patrick Cloke2022-10-051-17/+26
| | | | | | | | | | | | | | | | Fixes two related bugs: * The handling of `[null]` for a `room_types` filter was incorrect. * The ordering of arguments when providing both a network tuple and room type field was incorrect.
* | Use threaded receipts when fetching events for push. (#13878)Patrick Cloke2022-10-041-23/+57
| | | | | | | | Update the HTTP and email pushers to consider threaded read receipts when fetching unread events.
* | Remove get rooms for user with stream ordering (#13991)Nick Mills-Barrett2022-10-041-80/+69
| | | | | | | | | | | | By getting the joined rooms before the current token we avoid any reading history to confirm a user *was* in a room. We can then use any membership change events, which we already fetch during sync, to determine the final list of joined room IDs.
* | Recursively fetch the thread for receipts & notifications. (#13824)Patrick Cloke2022-10-043-2/+61
| | | | | | | | | | | | Consider an event to be part of a thread if you can follow a chain of relations up to a thread root. Part of MSC3773 & MSC3771.
* | Mark events as read using threaded read receipts from MSC3771. (#13877)Patrick Cloke2022-10-043-61/+315
| | | | | | | | Applies the proper logic for unthreaded and threaded receipts to either apply to all events in the room or only events in the same thread, respectively.
* | Advertise supporting version 1.3 of the Matrix spec. (#14032)Patrick Cloke2022-10-041-0/+1
| | | | | | Now that all features / changes in 1.3 are supported in Synapse.
* | Track notification counts per thread (implement MSC3773). (#13776)Patrick Cloke2022-10-0414-87/+333
| | | | | | | | | | | | | | | | When retrieving counts of notifications segment the results based on the thread ID, but choose whether to return them as individual threads or as a single summed field by letting the client opt-in via a sync flag. The summarization code is also updated to be per thread, instead of per room.
* | Disable pushing for server ACL events (MSC3786). (#13997)Patrick Cloke2022-10-042-10/+2
| | | | | | | | | | | | Switches to the stable identifier for MSC3786 and enables it by default. This disables pushes of m.room.server_acl events.
* | Send the appservice access token as a header. (#13996)Patrick Cloke2022-10-041-4/+19
|/ | | | | | | | | Implements MSC2832 by sending application service access tokens in the Authorization header. The access token is also still sent as a query parameter until the application service ecosystem has fully migrated to using headers. In the future this could be made opt-in, or removed completely.
* Track when the pulled event signature fails (#13815)Eric Eastwood2022-10-032-13/+62
| | | | | | | | | Because we're doing the recording in `_check_sigs_and_hash_for_pulled_events_and_fetch` (previously named `_check_sigs_and_hash_and_fetch`), this means we will track signature failures for `backfill`, `get_room_state`, `get_event_auth`, and `get_missing_events` (all pulled event scenarios). And we also record signature failures from `get_pdu`. Part of https://github.com/matrix-org/synapse/issues/13700 Part of https://github.com/matrix-org/synapse/issues/13676 and https://github.com/matrix-org/synapse/issues/13356 This PR will be especially important for https://github.com/matrix-org/synapse/pull/13816 so we can avoid the costly `_get_state_ids_after_missing_prev_event` down the line when `/messages` calls backfill.
* Clear out old rows from `event_push_actions_staging` (#14020)Erik Johnston2022-10-034-1/+104
| | | On matrix.org we have ~5 million stale rows in `event_push_actions_staging`, let's add a background job to make sure we clear them out.
* Announce that legacy metric names are deprecated, will be turned off by ↵reivilibre2022-10-031-26/+0
| | | | default in Synapse v1.71.0 and removed altogether in Synapse v1.73.0. (#14024)
* Do not return unspecced original_event field when using the stable ↵Patrick Cloke2022-10-032-12/+19
| | | | | | | | | | /relations endpoint. (#14025) Keep the old behavior (of including the original_event field) for any requests to the /unstable version of the endpoint, but do not include the field when the /v1 version is used. This should avoid new clients from depending on this field, but will not help with current dependencies.
* Fix bug where we didn't delete staging push actions (#14014)Erik Johnston2022-10-031-1/+1
| | | Introduced in #13719
* Add query parameter `ts` to allow appservices set the `origin_server_ts` for ↵lukasdenk2022-10-032-13/+34
| | | | | | | state events. (#11866) MSC3316 declares that both /rooms/{roomId}/send and /rooms/{roomId}/state should accept a ts parameter for appservices. This change expands support to /state and adds tests.
* Fix twisted trunk mypy errors (#14012)David Robertson2022-10-032-0/+6
|
* Add cache to `get_partial_state_servers_at_join` (#14013)Erik Johnston2022-10-031-0/+7
|
* Refactor `_get_e2e_device_keys_txn` to split large queries (#13956)Sean Quah2022-10-032-29/+114
| | | | | | | | | | Instead of running a single large query, run a single query for user-only lookups and additional queries for batches of user device lookups. Resolves #13580. Signed-off-by: Sean Quah <seanq@matrix.org>
* Revert the general exception recording introduced in #13814 (#13969)Eric Eastwood2022-10-031-10/+0
| | | | | | | | | * Maybe not catch all errors to avoid things in the nature-of CancelledError See https://github.com/matrix-org/synapse/pull/13815#discussion_r983384698 * Remove general exception tracking * Add changelog
* Fix `get_users_in_room` mis-use in `transfer_room_state_on_room_upgrade` ↵Eric Eastwood2022-09-301-2/+2
| | | | | | | | | | | | | | (#13960) Spawning from looking into `get_users_in_room` while investigating https://github.com/matrix-org/synapse/issues/13942#issuecomment-1262787050. See https://github.com/matrix-org/synapse/pull/13575#discussion_r953023755 for the original exploration around finding `get_users_in_room` mis-uses. Related to the following PRs where we also cleaned up some `get_users_in_room` mis-uses: - https://github.com/matrix-org/synapse/pull/13605 - https://github.com/matrix-org/synapse/pull/13608 - https://github.com/matrix-org/synapse/pull/13606 - https://github.com/matrix-org/synapse/pull/13958
* Clarifications in user directory for users who share rooms tracking (#13966)Eric Eastwood2022-09-301-12/+24
| | | | | Spawned while working on [`get_users_in_room` mis-uses](https://github.com/matrix-org/synapse/pull/13958#discussion_r984074897) and thinking we could use `get_local_users_in_room` here but we can't. From first glance, it seemed like this was only using local users from all of the `is_mine_id(user_id)` checks but I see that it does actually use remote users. Just making things a little more clear here what it does and mentions remote users so maybe that will be more obvious in the future.
* Skip filtering during push if there are no push actions (#13992)Erik Johnston2022-09-302-0/+9
|
* Update mypy and mypy-zope, attempt 3 (#13993)David Robertson2022-09-305-32/+20
| | | Co-authored-by: Patrick Cloke <clokep@users.noreply.github.com>
* Revert "Update mypy and mypy-zope (#13925)"David Robertson2022-09-305-20/+32
| | | | This reverts commit 6d543d6d9f56e39199b7e460d0081b02d61f12be.
* Update mypy and mypy-zope (#13925)David Robertson2022-09-305-32/+20
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | * Update mypy and mypy-zope * Unignore assigning to LogRecord attributes Presumably https://github.com/python/typeshed/pull/8064 makes this ok Cherry-picked from #13521 * Remove unused ignores due to mypy ParamSpec fixes https://github.com/python/mypy/pull/12668 Cherry-picked from #13521 * Remove additional unused ignores * Fix new mypy complaints related to `assertGreater` Presumably due to https://github.com/python/typeshed/pull/8077 * Changelog * Reword changelog Co-authored-by: Patrick Cloke <clokep@users.noreply.github.com> Co-authored-by: Patrick Cloke <clokep@users.noreply.github.com>
* Speed up calculating push actions in large rooms (#13973)Erik Johnston2022-09-301-10/+15
| | | | | We move the expensive check of visibility to after calculating push actions, avoiding the expensive check for users who won't get pushed anyway. I think this should have a big impact on rooms with large numbers of local users that have pushed disabled.
* Discourage automatic replies to Synapse's emails (#13957)David Robertson2022-09-301-0/+13
| | | Co-authored-by: Patrick Cloke <clokep@users.noreply.github.com>
* Fix performance regression in `get_users_in_room` (#13972)Erik Johnston2022-09-304-70/+97
| | | | | 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.
* Fix overflows in /messages backfill calculation (#13936)David Robertson2022-09-301-29/+53
| | | | | | | | * Reproduce bug * Compute `least_function` first * Substitute `least_function` with an f-string * Bugfix: avoid overflow Co-authored-by: Eric Eastwood <erice@element.io>
* Add missing version information in the ModuleApi (#13947)Brendan Abolivier2022-09-301-0/+4
|
* Don't require `setuptools_rust` at runtime (#13952)David Robertson2022-09-291-1/+16
|
* Update UPSERT comment now that native upserts are the default (#13924)David Robertson2022-09-291-10/+50
|
* Implement push rule evaluation in Rust. (#13838)Erik Johnston2022-09-293-381/+63
|
* Optimise get_rooms_for_user (drop with_stream_ordering) (#13787)Nick Mills-Barrett2022-09-295-75/+64
|
* Allow admins to require a manual approval process before new accounts can be ↵Brendan Abolivier2022-09-2914-26/+332
| | | | used (using MSC3866) (#13556)
* Explicit cast to enforce type hints. (#13939)Patrick Cloke2022-09-291-4/+4
|
* Expose MSC3882 only be under an unstable endpoint. (#13868)Hugh Nimmo-Smith2022-09-291-1/+3
|
* Clarify that a method returns only unthreaded receipts. (#13937)Patrick Cloke2022-09-292-40/+8
| | | | | By renaming it and updating the docstring. Additionally, refactors a method which is used only by tests.
* Always send default and rule_id to clients (#13904)Nicolas Werner2022-09-291-4/+2
|
* Improve backfill robustness by trying more servers. (#13890)reivilibre2022-09-291-2/+31
| | | Co-authored-by: Eric Eastwood <erice@element.io>
* Handle local device list updates during partial join (#13934)Erik Johnston2022-09-283-15/+140
|
* Limit and filter the number of backfill points to get from the database (#13879)Eric Eastwood2022-09-282-60/+139
| | | | | | | | | There is no need to grab thousands of backfill points when we only need 5 to make the `/backfill` request with. We need to grab a few extra in case the first few aren't visible in the history. Previously, we grabbed thousands of backfill points from the database, then sorted and filtered them in the app. Fetching the 4.6k backfill points for `#matrix:matrix.org` from the database takes ~50ms - ~570ms so it's not like this saves a lot of time 🤷. But it might save us more time now that `get_backfill_points_in_room`/`get_insertion_event_backward_extremities_in_room` are more complicated after https://github.com/matrix-org/synapse/pull/13635 This PR moves the filtering and limiting to the SQL query so we just have less data to work with in the first place. Part of https://github.com/matrix-org/synapse/issues/13356
* Revert "Stop returning an unused column when handling new receipts. ↵Patrick Cloke2022-09-281-2/+2
| | | | | | | (#13933)" (#13935) This reverts commit 7766bd5b354cd4ea1a33351ba320e54a14d3aeac (#13933). The unused column is actually used, but much further down in the function.
* Stop returning an unused column when handling new receipts. (#13933)Patrick Cloke2022-09-281-2/+2
|
* Handle remote device list updates during partial join (#13913)Erik Johnston2022-09-285-0/+169
| | | | | | | c.f. #12993 (comment), point 3 This stores all device list updates that we receive while partial joins are ongoing, and processes them once we have the full state. Note: We don't actually process the device lists in the same ways as if we weren't partially joined. Instead of updating the device list remote cache, we simply notify local users that a change in the remote user's devices has happened. I think this is safe as if the local user requests the keys for the remote user and we don't have them we'll simply fetch them as normal.
* fix: Push notifications for invite over federation (#13719)Kateřina Churanová2022-09-287-23/+41
|
* Persist CreateRoom events to DB in a batch (#13800)Shay2022-09-287-333/+542
|
* Prepatory work for batching events to send (#13487)Shay2022-09-283-106/+287
| | | 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.
* Fix `have_seen_event` cache not being invalidated (#13863)Eric Eastwood2022-09-272-18/+28
| | | | | | | | | | | | | | | | | | | | | Fix https://github.com/matrix-org/synapse/issues/13856 Fix https://github.com/matrix-org/synapse/issues/13865 > Discovered while trying to make Synapse fast enough for [this MSC2716 test for importing many batches](https://github.com/matrix-org/complement/pull/214#discussion_r741678240). As an example, disabling the `have_seen_event` cache saves 10 seconds for each `/messages` request in that MSC2716 Complement test because we're not making as many federation requests for `/state` (speeding up `have_seen_event` itself is related to https://github.com/matrix-org/synapse/issues/13625) > > But this will also make `/messages` faster in general so we can include it in the [faster `/messages` milestone](https://github.com/matrix-org/synapse/milestone/11). > > *-- https://github.com/matrix-org/synapse/issues/13856* ### The problem `_invalidate_caches_for_event` doesn't run in monolith mode which means we never even tried to clear the `have_seen_event` and other caches. And even in worker mode, it only runs on the workers, not the master (AFAICT). Additionally there was bug with the key being wrong so `_invalidate_caches_for_event` never invalidates the `have_seen_event` cache even when it does run. Because we were using the `@cachedList` wrong, it was putting items in the cache under keys like `((room_id, event_id),)` with a `set` in a `set` (ex. `(('!TnCIJPKzdQdUlIyXdQ:test', '$Iu0eqEBN7qcyF1S9B3oNB3I91v2o5YOgRNPwi_78s-k'),)`) and we we're trying to invalidate with just `(room_id, event_id)` which did nothing.
* Add new columns tracking when we partial-joined (#13892)David Robertson2022-09-273-3/+86
|
* Support the stable dir parameter for /relations. (#13920)Patrick Cloke2022-09-271-9/+15
| | | | | | | Since MSC3715 has passed FCP, the stable parameter can be used. This currently falls back to the unstable parameter if the stable parameter is not provided (and MSC3715 support is enabled in the configuration).
* Prioritize outbound to-device over device list updates (#13922)Erik Johnston2022-09-271-13/+16
| | | Otherwise device list changes for large accounts can temporarily delay to-device messages.
* Carry IdP Session IDs through user-mapping sessions. (#13839)Quentin Gliech2022-09-271-0/+9
| | | Since #11482, we're saving sessions IDs from upstream IdPs, but we've been losing them when the user goes through a user mapping session on account registration.
* Handle the case of remote users leaving a partial join room for device lists ↵Erik Johnston2022-09-276-107/+84
| | | | (#13885)
* Faster room joins: Fix spurious error when joining a room (#13872)Sean Quah2022-09-271-3/+19
| | | | | | | | | | | | | | | | During a `lazy_load_members` `/sync`, we look through auth events in rooms with partial state to find prior membership events. When such a membership is not found, an error is logged. Since the first join event for a user never has a prior membership event to cite, the error would always be logged when one appeared in the room timeline. Avoid logging errors for such events. Introduced in #13477. Signed-off-by: Sean Quah <seanq@matrix.org>
* Complement image: propagate SIGTERM to all workers (#13914)Richard van der Hoff2022-09-261-2/+30
| | | | | | | | | | This should mean that logs from worker processes are flushed before shutdown. When a test completes, Complement stops the docker container, which means that synapse will receive a SIGTERM. Currently, the `complement_fork_starter` exits immediately (without notifying the worker processes), which means that the workers never get a chance to flush their logs before the whole container is vaped. We can fix this by propagating the SIGTERM to the children.
* Improve tests for get_unread_push_actions_for_user_in_range_*. (#13893)Patrick Cloke2022-09-261-14/+24
| | | | | * Adds a docstring. * Reduces a small amount of duplicated code. * Improves tests.
* Snapshot schema 72 (#13873)David Robertson2022-09-2611-8/+2118
| | | Including another batch of fixes to the schema dump script
* typing: check origin server of typing event against room's servers (#13830)Mathieu Velten2022-09-261-2/+5
| | | | | | | This is also using the partial state approximation if needed so we do not block here during a fast join. Signed-off-by: Mathieu Velten <mathieuv@matrix.org> Co-authored-by: Sean Quah <8349537+squahtx@users.noreply.github.com>
* Simplify cache invalidation after event persist txn (#13796)Nick Mills-Barrett2022-09-263-119/+51
| | | | | This moves all the invalidations into a single place and de-duplicates the code involved in invalidating caches for a given event by using the base class method.
* Fix mypy errors with latest canonicaljson (#13905)David Robertson2022-09-261-2/+2
| | | | | | | | | | | | | | | | * Lockfile: update canonicaljson 1.6.0 -> 1.6.3 * Fix mypy errors with latest canonicaljson The change to `_encode_json_bytes` definition wasn't sufficient: ``` synapse/http/server.py:751: error: Incompatible types in assignment (expression has type "Callable[[Arg(object, 'json_object')], bytes]", variable has type "Callable[[Arg(object, 'data')], bytes]") [assignment] ``` Which I think is mypy warning us that the two functions accept different sets of kwargs. Fair enough! * Changelog
* Only try to backfill event if we haven't tried before recently (#13635)Eric Eastwood2022-09-232-44/+148
| | | | | | | | | | Only try to backfill event if we haven't tried before recently (exponential backoff). No need to keep trying the same backfill point that fails over and over. Fix https://github.com/matrix-org/synapse/issues/13622 Fix https://github.com/matrix-org/synapse/issues/8451 Follow-up to https://github.com/matrix-org/synapse/pull/13589 Part of https://github.com/matrix-org/synapse/issues/13356
* Faster room joins: Avoid blocking `/keys/changes` (#13888)Sean Quah2022-09-232-3/+11
| | | | | | | | | Part of the work for #12993. Once #12993 is fully resolved, we expect `/keys/changes` to behave sensibly when joined to a room with partial state. Signed-off-by: Sean Quah <seanq@matrix.org>
* Fix access token leak to logs from proxyagent (#13855)Eric Eastwood2022-09-231-1/+6
| | | | | | | | | | | | | | | This can happen specifically with an application service `/transactions/10722?access_token=leaked` request Fix https://github.com/matrix-org/synapse/issues/13010 --- Saw an example leak in https://github.com/matrix-org/synapse/issues/13423#issuecomment-1205348482 ``` 2022-08-04 14:47:57,925 - synapse.http.client - 401 - DEBUG - as-sender-signal-1 - Sending request PUT http://localhost:29328/transactions/10722?access_token=<redacted> 2022-08-04 14:47:57,926 - synapse.http.proxyagent - 223 - DEBUG - as-sender-signal-1 - Requesting b'http://localhost:29328/transactions/10722?access_token=leaked' via <HostnameEndpoint localhost:29328> ```
* Accept & store thread IDs for receipts (implement MSC3771). (#13782)Patrick Cloke2022-09-2310-27/+110
| | | | Updates the `/receipts` endpoint and receipt EDU handler to parse a `thread_id` from the body and insert it in the database.
* Send device list updates out to servers in partially joined rooms (#13874)Sean Quah2022-09-233-2/+65
| | | | | | | | | | | Use the provided list of servers in the room from the `/send_join` response, since we will not know which users are in the room. This isn't sufficient to ensure that all remote servers receive the right device list updates, since the `/send_join` response may be inaccurate or we may calculate the membership state of new users in the room incorrectly. Signed-off-by: Sean Quah <seanq@matrix.org>
* Faster Remote Room Joins: tell remote homeservers that we are unable to ↵reivilibre2022-09-239-41/+56
| | | | authorise them if they query a room which has partial state on our server. (#13823)
* Properly paginate forward in the /relations API. (#13840)Patrick Cloke2022-09-222-13/+31
| | | | | This fixes a bug where the `/relations` API with `dir=f` would skip the first item of each page (except the first page), causing incomplete data to be returned to the client.
* Last batch of Pydantic for synapse/rest/client/account.py (#13832)David Robertson2022-09-211-6/+13
| | | | | | | * Validation for `/add_threepid/msisdn/submit_token` * Don't validate deprecated endpoint * Changelog
* Add version flag for MSC3881 (#13860)Brendan Abolivier2022-09-211-0/+2
|
* Track device IDs for pushers (#13831)Brendan Abolivier2022-09-215-5/+103
| | | Second half of the MSC3881 implementation
* Implementation of MSC3882 login token request (#13722)Hugh Nimmo-Smith2022-09-214-0/+105
|
* Support enabling/disabling pushers (from MSC3881) (#13799)Brendan Abolivier2022-09-2110-54/+154
| | | Partial implementation of MSC3881
* Add cache invalidation across workers to module API (#13667)Mathieu Velten2022-09-214-19/+71
| | | Signed-off-by: Mathieu Velten <mathieuv@matrix.org>
* Correct documentation for map_user_attributes of OpenID Mapping Providers ↵Peter Scheu2022-09-211-0/+3
| | | | | (#13836) Co-authored-by: David Robertson <davidr@element.io>
* Remove the `complete_sso_login` method from the Module API which was ↵Quentin Gliech2022-09-202-58/+1
| | | | | deprecated in Synapse 1.13.0. (#13843) Signed-off-by: Quentin Gliech <quenting@element.io>
* Generate separate snapshots for logical databases (#13792)David Robertson2022-09-202-5/+14
| | | | | | | * Generate separate snapshots for sqlite, postgres and common * Cleanup postgres dbs in the TRAP * Say which logical DB we're applying updates to * Run background updates on the state DB * Add new option for accepting a SCHEMA_NUMBER
* Port the push rule classes to Rust. (#13768)Erik Johnston2022-09-205-598/+25
|
* Don't include redundant prev_state in new events (#13791)Denis2022-09-202-4/+0
|
* Add support to purge rows from MSC2716 and other tables when purging a room ↵Eric Eastwood2022-09-163-0/+29
| | | | | | | | | | | (#13825) `event_failed_pull_attempts` added in https://github.com/matrix-org/synapse/pull/13589 MSC2716 related tables added in: - https://github.com/matrix-org/synapse/pull/10245/files#diff-3d42dfb44d02f7de3aada105e0bdc1cc9dd7f953cbf0f36c5d0f50827bf0320aR1 - Renamed in https://github.com/matrix-org/synapse/pull/10838/files#diff-2730bfbe9e688b55e46f9371aefe67dac2bd2b2b7d9d6b92774eea1fcfae156dR1 - https://github.com/matrix-org/synapse/pull/10498/files#diff-c52bbfbb5921a3f6f023b24343668479d966fac164f13b7c39d2197ce3afa7a5R1
* Remove error spam when users query the keys of departed remote users (#13826)Sean Quah2022-09-161-9/+12
| | | | The error message introduced in #13749 has turned out to be very spammy. Remove it for now.
* Add an admin API endpoint to find a user based on its external ID in an auth ↵Quentin Gliech2022-09-162-0/+29
| | | | provider. (#13810)
* Avoid putting rejected events in room state (#13723)Sean Quah2022-09-161-0/+15
| | | Signed-off-by: Sean Quah <seanq@matrix.org>
* Be able to correlate timeouts in reverse-proxy layer in front of Synapse ↵Eric Eastwood2022-09-152-4/+23
| | | | | | | | | | | | | | | | | | (pull request ID from header) (#13801) Fix https://github.com/matrix-org/synapse/issues/13685 New config: ```diff listeners: - port: 8008 tls: false type: http x_forwarded: true + request_id_header: "cf-ray" bind_addresses: ['::1', '127.0.0.1', '0.0.0.0'] ```
* Record any exception when processing a pulled event (#13814)Eric Eastwood2022-09-151-0/+10
| | | | | Part of https://github.com/matrix-org/synapse/issues/13700 and https://github.com/matrix-org/synapse/issues/13356 Follow-up to https://github.com/matrix-org/synapse/pull/13589
* Support providing an index predicate for upserts. (#13822)Patrick Cloke2022-09-152-7/+24
| | | | This is useful to upsert against a table which has a unique partial index while avoiding conflicts.
* A third batch of Pydantic validation for rest/client/account.py (#13736)David Robertson2022-09-152-42/+51
|
* Add a `MXCUri` class to make working with mxc uri's easier. (#13162)Andrew Morgan2022-09-152-4/+8
|
* Keep track when we try and fail to process a pulled event (#13589)Eric Eastwood2022-09-145-9/+106
| | | | | | | | | | | | | | We can follow-up this PR with: 1. Only try to backfill from an event if we haven't tried recently -> https://github.com/matrix-org/synapse/issues/13622 1. When we decide to backfill that event again, process it in the background so it doesn't block and make `/messages` slow when we know it will probably fail again -> https://github.com/matrix-org/synapse/issues/13623 1. Generally track failures everywhere we try and fail to pull an event over federation -> https://github.com/matrix-org/synapse/issues/13700 Fix https://github.com/matrix-org/synapse/issues/13621 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.qv7cj51sv9i5)
* Update event push action and receipt tables to support threads. (#13753)Patrick Cloke2022-09-149-20/+310
| | | | | | | | | | | | | | | Adds a `thread_id` column to the `event_push_actions`, `event_push_actions_staging`, and `event_push_summary` tables. This will notifications to be segmented by the thread in a future pull request. The `thread_id` column stores the root event ID or the special value `"main"`. The `thread_id` column for `event_push_actions` and `event_push_summary` is backfilled with `"main"` for all existing rows. New entries into `event_push_actions` and `event_push_actions_staging` will get the proper thread ID. `receipts_linearized` and `receipts_graph` also gain a `thread_id` column, which is similar, except `NULL` is a special value meaning the receipt is "unthreaded". See MSC3771 and MSC3773 for where this data will be useful.
* Use partial indices on SQLIte. (#13802)Patrick Cloke2022-09-143-5/+58
| | | | | | | Partial indices have been supported since SQLite 3.8, but Synapse now requires >= 3.27, so we can enable support for them. This requires rebuilding previous indices which were partial on PostgreSQL, but not on SQLite.
* Deduplicate `is_server_notices_room`. (#13780)reivilibre2022-09-143-18/+19
|
* Fix a memory leak when running the unit tests. (#13798)reivilibre2022-09-142-6/+7
|
* Remove unused method in `synapse.api.auth.Auth`. (#13795)Quentin Gliech2022-09-141-9/+0
| | | | | Clean-up from b19060a29b4f73897847db2aba5d03ec819086e0 (#13094) and 73af10f419346a5f2d70131ac1ed8e69942edca0 (#13093) which removed all callers.
* Remove incorrect migration file from `state` logical DB (#13788)David Robertson2022-09-141-37/+0
| | | | | | | | | | | | | * Remove incorrect migration file from `state` logical DB The table `ex_outlier_stream` is part of the `main` logical DB; it should not have been created in the `state` logical DB. We remove this migration now as a tidy-up. Note: we cannot `DROP TABLE IF EXISTS ex_outlier_stream` in a new migration, because some (most) instances of Synapse host both of these logical DBs on the same DB cluster. * Changelog
* Fix bug in device list caching when remote users leave rooms (#13749)Sean Quah2022-09-143-14/+43
| | | | | | | | | | | | When a remote user leaves the last room shared with the homeserver, we have to mark their device list as unsubscribed, otherwise we would hold on to a stale device list in our cache. Crucially, the device list would remain cached even after the remote user rejoined the room, which could lead to E2EE failures until the next change to the remote user's device list. Fixes #13651. Signed-off-by: Sean Quah <seanq@matrix.org>
* Fix a long-standing spec compliance bug where Synapse would accept a ↵reivilibre2022-09-141-2/+1
| | | | | | | | | | | trailing slash on the end of `/get_missing_events` federation requests. (#13789) * Don't accept a trailing slash on the end of /get_missing_events * Newsfile Signed-off-by: Olivier Wilkinson (reivilibre) <oliverw@matrix.org> Signed-off-by: Olivier Wilkinson (reivilibre) <oliverw@matrix.org>
* Make sequence `cache_invalidation_stream_seq` begin at `2` (#13766)Mathieu Velten2022-09-132-0/+24
| | | | Signed-off-by: Mathieu Velten <mathieuv@matrix.org> Co-authored-by: Sean Quah <8349537+squahtx@users.noreply.github.com>
* Updates to the schema dump script (#13770)David Robertson2022-09-131-0/+4
|
* Add receipts event stream ordering (#13703)Nick Mills-Barrett2022-09-133-1/+94
|
* Remove check current state membership up to date (#13745)Nick Mills-Barrett2022-09-122-155/+99
| | | | | | | * Remove checks for membership column in current_state_events * Add schema script to force through the `current_state_events_membership` background job Contributed by Nick @ Beeper (@fizzadar).
* Check if Rust lib needs rebuilding. (#13759)Erik Johnston2022-09-122-0/+89
| | | This protects against the common mistake of failing to remember to rebuild Rust code after making changes.
* Concurrently collect room unread counts for push badges (#13765)Nick Mills-Barrett2022-09-091-3/+10
| | | | | | | Most of the time this function is heavily cached, but when that isn't the case fetching the counts room by room slows down push delivery on users with many (thousands) of rooms. Signed off by Nick @ Beeper.
* Tag trace with instance name (#13761)Eric Eastwood2022-09-092-2/+11
| | | | | | | | We tag the Synapse instance name so that it's an easy jumping off point into the logs. Can also be used to filter for an instance that is under load. As suggested by @clokep and @reivilibre in, - https://github.com/matrix-org/synapse/pull/13729#discussion_r964719258 - https://github.com/matrix-org/synapse/pull/13729#discussion_r964733578
* Strip number suffix from instance name to consolidate services that traces ↵Eric Eastwood2022-09-091-1/+12
| | | | | | | | | | | | are spread over (#13729) The problem with many services is that it makes it hard to find which service has the trace you want, see https://github.com/jaegertracing/jaeger-ui/issues/985 Previously, we split traces out into services based on their instance name like `matrix.org client_reader-1`, etc but there are many worker instances of the same `client_reader` so there is a lot to click through. With this PR, all of the traces are just collected under the worker type like `client_reader`, `event_persister` 😇 Note: A Synapse worker instance name is an opaque string with the number convention only being our own thing for the `matrix.org` deployment. But seems pretty sensible to group things this way.
* Use an upsert for `receipts_graph`. (#13752)Patrick Cloke2022-09-091-8/+4
| | | | | | Instead of a delete, then insert. This was previously done for `receipts_linearized` in 2dc430d36ef793b38d6d79ec8db4ea60588df2ee (#7607).
* Require SQLite >= 3.27.0 (#13760)David Robertson2022-09-097-207/+105
|
* Re-type hint some collections in `/sync` code as read-only (#13754)Sean Quah2022-09-081-10/+10
| | | | Signed-off-by: Sean Quah <seanq@matrix.org>
* Add timestamp to user's consent (#13741)Dirk Klimpel2022-09-083-1/+22
| | | Co-authored-by: reivilibre <olivier@librepush.net>
* Update docstrings to explain the impact of partial state (#13750)Sean Quah2022-09-081-1/+16
| | | | | | | Update the docstrings for `get_users_in_room` and `get_current_hosts_in_room` to explain the impact of partial state. Signed-off-by: Sean Quah <seanq@matrix.org>
* Avoid raising errors due to malformed IDs in `get_current_hosts_in_room` ↵Sean Quah2022-09-081-1/+4
| | | | | | | | | | (#13748) Handle malformed user IDs with no colons in `get_current_hosts_in_room`. It's not currently possible for a malformed user ID to join a room, so this error would never be hit. Signed-off-by: Sean Quah <seanq@matrix.org>
* Fix error in `is_mine_id` when encountering a malformed ID (#13746)Sean Quah2022-09-081-1/+11
| | | | | | | | | Previously, `is_mine_id` would raise an exception when passed an ID with no colons. Return `False` instead. Fixes #13040. Signed-off-by: Sean Quah <seanq@matrix.org>
* Fix cache metrics not being updated when not using the legacy exposition ↵reivilibre2022-09-083-21/+80
| | | | module. (#13717)
* Fix Prometheus recording rules to not use legacy metric names. (#13718)reivilibre2022-09-083-5/+10
|
* Fix a bug where Synapse fails to start if a signing key file contains an ↵reivilibre2022-09-081-1/+12
| | | | empty line. (#13738)
* Instrument `get_metadata_for_events` for tracing (#13730)Eric Eastwood2022-09-071-0/+2
| | | | When backfilling, `_get_state_ids_after_missing_prev_event` calls [`get_metadata_for_events`](https://github.com/matrix-org/synapse/blob/26bc26586b4b95d63ce7e453e9312469843f796e/synapse/handlers/federation_event.py#L1133). For `#matrix:matrix.org`, it's called with 77k `state_events` which means 77 calls to the database and takes 28 seconds.
* A second batch of Pydantic models for rest/client/account.py (#13687)David Robertson2022-09-073-34/+63
|
* Cancel the processing of key query requests when they time out. (#13680)reivilibre2022-09-0715-19/+71
|
* Rename the `EventFormatVersions` enum values so that they line up with room ↵reivilibre2022-09-079-37/+42
| | | | version numbers. (#13706)
* Add Admin API to Fetch Messages Within a Particular Window (#13672)Connor Davis2022-09-073-13/+132
| | | This adds two new admin APIs that allow us to fetch messages from a room within a particular time.
* Remove the unspecced room_id field in the /hierarchy response. (#13506)reivilibre2022-09-061-1/+0
| | | | | | | | | | | This is a re-do of 57d334a13d983406ea452dfa203bbe4837509c4e (#13365), which was backed out in 12abd724974a2311d5311272d26d2f8aa11734a9 (#13501). The `room_id` field represented the parent space for each room and was made redundant by changes in the API shape where the `children_state` is now nested underneath each `room`. The room ID of each child is in the `state_key` field and is still available.
* Actually fix typechecking with latest types-jsonschema (#13724)David Robertson2022-09-061-4/+4
|
* Update Grafana dashboard to not use legacy metric names. (#13714)reivilibre2022-09-062-4/+4
|
* Remove configuration options for direct TCP replication. (#13647)Patrick Cloke2022-09-064-54/+39
| | | Removes the ability to configure legacy direct TCP replication. Workers now require Redis to run.
* Fix typechecking with latest `types-jsonschema` (#13712)David Robertson2022-09-051-4/+4
|
* Share some metrics between the Prometheus exporter and the phone home stats ↵Brendan Abolivier2022-09-054-3/+100
| | | | (#13671)
* Add a schema delta to drop unstable private read receipts. (#13692)Patrick Cloke2022-09-011-0/+19
| | | | Otherwise they'll be leaked due to the filtering code only respecting the stable identifiers for private read receipts.
* Disable calculating unread counts unless the config flag is enabled. (#13694)Patrick Cloke2022-09-012-1/+9
| | | | | | | | This avoids doing work that will never be used (since the resulting unread counts will never be sent in a /sync response). The negative of doing this is that unread counts will be incorrect when the feature is initially enabled.
* Cache `is_partial_state_room` (#13693)Erik Johnston2022-09-011-4/+7
| | | Fixes #13613.
* Add some logging to help track down #13444 (#13679)Erik Johnston2022-09-011-0/+13
|
* Return keys for unwhitelisted servers from `/_matrix/key/v2/query` (#13683)Richard van der Hoff2022-09-011-20/+21
|
* Remove support for unstable private read receipts (#13653)Šimon Brandner2022-09-019-39/+7
| | | Signed-off-by: Šimon Brandner <simon.bra.ag@gmail.com>
* Drop support for calling `/_matrix/client/v3/rooms/{roomId}/invite` without ↵Jacek Kuśnierz2022-08-315-134/+55
| | | | | | | an `id_access_token` (#13241) Fixes #13206 Signed-off-by: Jacek Kusnierz jacek.kusnierz@tum.de
* Remove cached wrap on `_get_joined_users_from_context` method (#13569)Nick Mills-Barrett2022-08-312-85/+39
| | | | | | | The method doesn't actually do any data fetching and the method that does, `_get_joined_profile_from_event_id`, has its own cache. Signed off by Nick @ Beeper (@Fizzadar).
* Generalise the `@cancellable` annotation so it can be used on functions ↵reivilibre2022-08-315-71/+68
| | | | other than just servlet methods. (#13662)
* Fix admin List Room API return type on sqlite (#13509)David Robertson2022-08-311-2/+4
|
* Give the correct next event when the message timestamps are the same - ↵Eric Eastwood2022-08-301-2/+10
| | | | | | | | | MSC3030 (#13658) Discovered while working on https://github.com/matrix-org/synapse/pull/13589 and I had all the messages at the same timestamp in the tests. Part of https://github.com/matrix-org/matrix-spec-proposals/pull/3030 Complement tests: https://github.com/matrix-org/complement/pull/457
* Drop unused column `application_services_state.last_txn` (#13627)Shay2022-08-303-0/+58
|
* Merge branch 'release-v1.66' into developDavid Robertson2022-08-302-32/+127
|\
| * Fix rate limit metrics registering twice and misreporting (#13649)Eric Eastwood2022-08-302-32/+127
| | | | | | | | | | | | | | | | | | | | | | * Fix rate limit metrics registering twice and misreporting Fix https://github.com/matrix-org/synapse/issues/13641 * Fix lints * Add changelog * Document `metrics_name=None`.
* | Fix bug where we wedge media plugins if clients disconnect early (#13660)Erik Johnston2022-08-301-19/+21
| | | | | | | | | | | | | | | | We incorrectly didn't use the returned `Responder` if the client had disconnected, which meant that the resource used by the Responder wasn't correctly released. In particular, this exhausted the thread pools so that *all* requests timed out.
* | Do not wait for background updates to complete do expire URL cache. (#13657)Patrick Cloke2022-08-301-4/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | Media downloaded as part of a URL preview is normally deleted after two days. However, while a background database migration is running, the process is stopped. A long-running database migration can therefore cause the media store to fill up with old preview files. This logic was added in #2697 to make sure that we didn't try to run the expiry without an index on `local_media_repository.created_ts`; the original logic that needs that index was added in #2478 (in `get_url_cache_media_before`, as amended by 93247a424a5068b088567fa98b6990e47608b7cb), and is still present. Given that the background update was added before Synapse v1.0.0, just drop this check and assume the index exists.
* | Speed up inserting `event_push_actions_staging`. (#13634)Patrick Cloke2022-08-301-20/+8
| | | | | | By using `execute_values` instead of `execute_batch`.
* | Fix that user cannot `/forget` rooms after the last member has left (#13546)Dirk Klimpel2022-08-301-2/+5
| |
* | Optimize how we calculate `likely_domains` during backfill (#13575)Eric Eastwood2022-08-304-70/+88
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 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/
* | Generate missing configuration files at startup (#13615)Richard van der Hoff2022-08-261-11/+48
| | | | | | | | | | | | | | | | If things like the signing key file are missing, let's just try to generate them on startup. Again, this is useful for k8s-like deployments where we just want to generate keys on the first run.
* | Move the execution of the retention purge_jobs to the main worker (#13632)Brad Murray2022-08-261-4/+2
| | | | | | | | | | Fixes #9927 Signed-off-by: Brad Murray brad@beeper.com
* | Support `registration_shared_secret` in a file (#13614)Richard van der Hoff2022-08-252-5/+73
| | | | | | | | A new `registration_shared_secret_path` option. This is kinda handy for k8s deployments and things.
* | register_new_matrix_user: read server url from config (#13616)Richard van der Hoff2022-08-251-6/+51
| | | | | | | | Fixes https://github.com/matrix-org/synapse/issues/3672: `https://localhost:8448` is virtually never right.
generated by cgit-pink 1.4.1 (git 2.36.1) at 2024-11-18 03:51:58 +0000