Commit Graph

12549 Commits

Author SHA1 Message Date
Will Hunt
626b8f0846
Move schema file for as_device_stream (#8590)
* Move schema file

* Add a .

* Add matching changelog entry

* Fix sqlite
2020-10-20 10:18:55 +01:00
Richard van der Hoff
96e7d3c4a0
Fix 'LruCache' object has no attribute '_on_resize' (#8591)
We need to make sure we are readu for the `set_cache_factor` callback.
2020-10-19 21:13:50 +01:00
Vasilis Gerakaris
34c20493b9
Drop unused device_max_stream_id table (#8589)
Signed-off-by: Vasilis Gerakaris <vasilis.gerakaris@navarino.gr>
2020-10-19 19:06:54 +01:00
Jonathan de Jong
21bb50ca3f
Fix mypy error: auth handler "checkpw" internal function type mismatch (#8569) 2020-10-19 18:32:24 +01:00
Patrick Cloke
8f27b7fde1
Expose the experimental appservice login flow to clients. (#8504) 2020-10-19 13:03:55 -04:00
Richard van der Hoff
903d11c43a
Add DeferredCache.get_immediate method (#8568)
* Add `DeferredCache.get_immediate` method

A bunch of things that are currently calling `DeferredCache.get` are only
really interested in the result if it's completed. We can optimise and simplify
this case.

* Remove unused 'default' parameter to DeferredCache.get()

* another get_immediate instance
2020-10-19 15:00:12 +01:00
Patrick Cloke
c356b4bf42
Include a simple message in email notifications that include encrypted content (#8545) 2020-10-19 09:12:39 -04:00
Erik Johnston
1fcdbeb3ab
Start an opentracing span for background processes. (#8567)
This should reduce the number of `There was no active span` errors we
see.

Fixes #8510.
2020-10-19 12:26:26 +01:00
Richard van der Hoff
97647b33c2
Replace DeferredCache with LruCache where possible (#8563)
Most of these uses don't need a full-blown DeferredCache; LruCache is lighter and more appropriate.
2020-10-19 12:20:29 +01:00
Jonathan de Jong
79c1f973ce
Pre-emptively fix synapse.storage.types.Connection for future mypy release (#8577)
Fix the Connection protocol according to typeshed's assertions about sqlite3.Connection
2020-10-17 09:51:38 +01:00
Richard van der Hoff
d6094176d1
Type annotations for LruCache (#8562)
* type annotations for LruCache

* changelog

* Apply suggestions from code review

Co-authored-by: Patrick Cloke <clokep@users.noreply.github.com>

* review comments

Co-authored-by: Patrick Cloke <clokep@users.noreply.github.com>
2020-10-16 17:06:50 +01:00
Patrick Cloke
1b70662be9
Clean-up old transaction IDs on the background worker. (#8544) 2020-10-16 12:06:17 -04:00
Richard van der Hoff
6d7b22041d review comments 2020-10-16 16:25:15 +01:00
Richard van der Hoff
995cc615a0
Apply suggestions from code review
Co-authored-by: Patrick Cloke <clokep@users.noreply.github.com>
2020-10-16 16:14:42 +01:00
Richard van der Hoff
0ec0bc3886 type annotations for LruCache 2020-10-16 15:56:39 +01:00
Richard van der Hoff
3ee17585cd
Make LruCache register its own metrics (#8561)
rather than have everything that instantiates an LruCache manage metrics
separately, have LruCache do it itself.
2020-10-16 15:51:57 +01:00
Richard van der Hoff
da0090fdff
Fix modifying events in ThirdPartyRules modules (#8564)
EventBuilder.build wants auth events these days
2020-10-16 13:39:46 +01:00
Richard van der Hoff
5649669c3c
Merge pull request #8535 from matrix-org/rav/third_party_events_updates
Support modifying event content from ThirdPartyRules modules
2020-10-15 20:55:41 +01:00
Jonathan de Jong
6b5a115c0a
Solidify the HomeServer constructor. (#8515)
This implements a more standard API for instantiating a homeserver and
moves some of the dependency injection into the test suite.

More concretely this stops using `setattr` on all `kwargs` passed to `HomeServer`.
2020-10-15 15:29:13 -04:00
Will Hunt
c276bd9969
Send some ephemeral events to appservices (#8437)
Optionally sends typing, presence, and read receipt information to appservices.
2020-10-15 12:33:28 -04:00
Patrick Cloke
74976a8e43 Merge branch 'master' into develop 2020-10-15 10:43:54 -04:00
Patrick Cloke
9991aaa49c 1.21.2 2020-10-15 09:24:10 -04:00
Erik Johnston
20fa83f374 Remove racey assertion in MultiWriterIDGenerator (#8530)
We asserted that the IDs returned by postgres sequence was greater than
any we had seen, however this is technically racey as we may update the
current positions out of order.

We now assert that the sequences are correct on startup, so the
assertion is no longer really required, so we remove them.
2020-10-15 09:14:29 -04:00
Richard van der Hoff
8075504a60
Enable mypy for synapse.util.caches (#8547)
This seemed to entail dragging in a type stub for SortedList.
2020-10-15 11:44:39 +01:00
Richard van der Hoff
0a08cd1065
Merge pull request #8548 from matrix-org/rav/deferred_cache
Rename Cache to DeferredCache, and related changes
2020-10-15 11:42:07 +01:00
Neil Johnson
1f39155071
Include user agent in user daily visits table (#8503)
Include user agent in user daily visits table.
2020-10-15 10:36:40 +01:00
Richard van der Hoff
4433d01519
Merge pull request #8537 from matrix-org/rav/simplify_locally_reject_invite
Simplify `_locally_reject_invite`
2020-10-15 10:20:19 +01:00
Richard van der Hoff
4182bb812f move DeferredCache into its own module 2020-10-14 23:38:14 +01:00
Richard van der Hoff
9f87da0a84 Rename Cache->DeferredCache 2020-10-14 23:38:14 +01:00
Richard van der Hoff
7eff59ec91 Add some more type annotations to Cache 2020-10-14 23:38:14 +01:00
Erik Johnston
19b15d63e8
Use autocommit mode for single statement DB functions. (#8542)
Autocommit means that we don't wrap the functions in transactions, and instead get executed directly. Introduced in #8456. This will help:

1. reduce the number of `could not serialize access due to concurrent delete` errors that we see (though there are a few functions that often cause serialization errors that we don't fix here);
2. improve the DB performance, as it no longer needs to deal with the overhead of `REPEATABLE READ` isolation levels; and
3. improve wall clock speed of these functions, as we no longer need to send `BEGIN` and `COMMIT` to the DB.

Some notes about the differences between autocommit mode and our default `REPEATABLE READ` transactions:

1. Currently `autocommit` only applies when using PostgreSQL, and is ignored when using SQLite (due to silliness with [Twisted DB classes](https://twistedmatrix.com/trac/ticket/9998)).
2. Autocommit functions may get retried on error, which means they can get applied *twice* (or more) to the DB (since they are not in a transaction the previous call would not get rolled back). This means that the functions need to be idempotent (or otherwise not care about being called multiple times). Read queries, simple deletes, and updates/upserts that replace rows (rather than generating new values from existing rows) are all idempotent.
3. Autocommit functions no longer get executed in [`REPEATABLE READ`](https://www.postgresql.org/docs/current/transaction-iso.html) isolation level, and so data can change queries, which is fine for single statement queries.
2020-10-14 15:50:59 +01:00
Erik Johnston
618d405a32
Remove racey assertion in MultiWriterIDGenerator (#8530)
We asserted that the IDs returned by postgres sequence was greater than
any we had seen, however this is technically racey as we may update the
current positions out of order.

We now assert that the sequences are correct on startup, so the
assertion is no longer really required, so we remove them.
2020-10-14 15:40:06 +01:00
Erik Johnston
921a3f8a59
Fix not sending events over federation when using sharded event persisters (#8536)
* Fix outbound federaion with multiple event persisters.

We incorrectly notified federation senders that the minimum persisted
stream position had advanced when we got an `RDATA` from an event
persister.

Notifying of federation senders already correctly happens in the
notifier, so we just delete the offending line.

* Change some interfaces to use RoomStreamToken.

By enforcing use of `RoomStreamTokens` we make it less likely that
people pass in random ints that they got from somewhere random.
2020-10-14 13:27:51 +01:00
Brendan Abolivier
3ee97a2748
Make sure a retention policy is a state event (#8527)
* Make sure a retention policy is a state event

* Changelog
2020-10-14 12:00:52 +01:00
Aaron Raimist
ec606ea9e3
Add correct M_BAD_JSON error code to /profile/{userId}/displayname (#8517)
Fixes #8029
2020-10-14 11:24:58 +01:00
Richard van der Hoff
a34b17e492 Simplify _locally_reject_invite
Update `EventCreationHandler.create_event` to accept an auth_events param, and
use it in `_locally_reject_invite` instead of reinventing the wheel.
2020-10-13 23:58:48 +01:00
Richard van der Hoff
898196f1cc guard against accidental modification 2020-10-13 23:24:50 +01:00
Richard van der Hoff
617e8a4653 Allow ThirdPartyRules modules to replace event content
Support returning a new event dict from `check_event_allowed`.
2020-10-13 23:24:50 +01:00
Richard van der Hoff
d9d86c2996 Remove redundant token_id parameter to create_event
this is always the same as requester.access_token_id.
2020-10-13 23:06:36 +01:00
Richard van der Hoff
123711ed19 Move third_party_rules check to event creation time
Rather than waiting until we handle the event, call the ThirdPartyRules check
when we fist create the event.
2020-10-13 21:38:48 +01:00
Richard van der Hoff
d59378d86b Remove redundant calls to third_party_rules in on_send_{join,leave}
There's not much point in calling these *after* we have decided to accept them
into the DAG.
2020-10-13 21:38:48 +01:00
Patrick Cloke
629a951b49
Move additional tasks to the background worker, part 4 (#8513) 2020-10-13 08:20:32 -04:00
Erik Johnston
b2486f6656
Fix message duplication if something goes wrong after persisting the event (#8476)
Should fix #3365.
2020-10-13 12:07:56 +01:00
Andrew Morgan
a9a8f29729 Merge branch 'master' into develop
* master:
  1.21.1
  Explicitly install test dependencies when building deb packages (#8523)
2020-10-13 10:33:58 +01:00
Andrew Morgan
58e583eac1 1.21.1 2020-10-13 10:27:16 +01:00
Andrew Morgan
bc203c962f Merge branch 'master' into develop 2020-10-12 16:41:52 +01:00
Erik Johnston
8de3703d21
Make event persisters periodically announce position over replication. (#8499)
Currently background proccesses stream the events stream use the "minimum persisted position" (i.e. `get_current_token()`) rather than the vector clock style tokens. This is broadly fine as it doesn't matter if the background processes lag a small amount. However, in extreme cases (i.e. SyTests) where we only write to one event persister the background processes will never make progress.

This PR changes it so that the `MultiWriterIDGenerator` keeps the current position of a given instance as up to date as possible (i.e using the latest token it sees if its not in the process of persisting anything), and then periodically announces that over replication. This then allows the "minimum persisted position" to advance, albeit with a small lag.
2020-10-12 15:51:41 +01:00
Andrew Morgan
f76194a021 1.21.0 2020-10-12 15:50:27 +01:00
Mateusz Przybyłowicz
ca2db5dd0c
Increase default max_upload_size from 10M to 50M (#8502)
Signed-off-by: Mateusz Przybyłowicz <uamfhq@gmail.com>
2020-10-09 16:58:23 +01:00
Patrick Cloke
1781bbe319
Add type hints to response cache. (#8507) 2020-10-09 11:35:11 -04:00