When we're adding state to the database, we check which eventNIDs are
already in a block, if we already have that eventNID, we remove it from
the list. In its current form we would skip over eventNIDs in the case
we already found a match (we're decrementing `i` twice)
My theory is, that when we later get the state blocks, we are receiving
"too many" eventNIDs (well, yea, we stored too many), which may or may
not can result in state resets when comparing different state snapshots.
(e.g. when adding state we stored a eventNID by accident because we
skipped it, later we add more state and are not adding it because we
don't skip it)
Requires https://github.com/matrix-org/gomatrixserverlib/pull/376
This has numerous upsides:
- Less type casting to `*Event` is required.
- Making Dendrite work with `PDU` interfaces means we can swap out Event
impls more easily.
- Tests which represent weird event shapes are easier to write.
Part of a series of refactors on GMSL.
We only use it in a few places currently, enough to get things to
compile and run. We should be using it in much more places.
Similarly, in some places we cast []PDU back to []*Event, we need to not
do that. Likewise, in some places we cast PDU to *Event, we need to not
do that. For now though, hopefully this is a start.
Replaced with types.HeaderedEvent _for now_. In reality we want to move
them all to gmsl.Event and only use HeaderedEvent when we _need_ to
bundle the version/event ID with the event (seriailsation boundaries,
and even then only when we don't have the room version).
Requires https://github.com/matrix-org/gomatrixserverlib/pull/373
Adds tests for `QueryRestrictedJoinAllowed`, `IsServerAllowed` and
`PerformRoomUpgrade`. Refactors the `QueryRoomVersionForRoom` method to
accept a string and return a `gmsl.RoomVersion` instead of req/resp
structs.
Adds some more caching for `GetStateEvent`
This should also fix#2912 by ignoring state events belonging to other
users.
As outlined in https://github.com/matrix-org/gomatrixserverlib/pull/368
The main change Dendrite side is that `RoomVersion` no longer has any
methods on it. Instead, you need to bounce via `gmsl.GetRoomVersion`.
It's very interesting to see where exactly Dendrite cares about this.
For some places it's creating events (fine) but others are way more
specific. Those areas will need to migrate to GMSL at some point.
This should deflake UTs and be more correct in terms of getting
`Events`.
`Events` tries to fetch the event from the cache first and may get an
unredacted event from it, while it should already be redacted.
We need to check the redaction PL in Dendrite, if we do it in GMSL, we
end up not sending the event to the output stream because it will be
rejected.
---------
Co-authored-by: kegsay <kegan@matrix.org>
This PR changes the following:
- `StoreEvent` now only stores an event (and possibly prev event),
instead of also doing redactions
- Adds a `MaybeRedactEvent` (pulled out from `StoreEvent`), which should
be called after storing events
- a few other things
This PR changes a few things:
- It pulls out the creation of several NIDs from the `StoreEvent`
function to make the functions more reusable
- Uses more caching when using those NIDs to avoid DB round trips
Should fix the following issues or make a lot less worse when using
Postgres:
The main issue behind #2911: The client gives up after a certain time,
causing a cascade of context errors, because the response couldn't be
built up fast enough. This mostly happens on accounts with many rooms,
due to the inefficient way we're getting recent events and current state
For #2777: The queries for getting the membership events for history
visibility were being executed for each room (I think 185?), resulting
in a whooping 2k queries for membership events. (Getting the
statesnapshot -> block nids -> actual wanted membership event)
Both should now be better by:
- Using a LATERAL join to get all recent events for all joined rooms in
one go (TODO: maybe do the same for room summary and current state etc)
- If we're lazy loading on initial syncs, we're now not getting the
whole current state, just to drop the majority of it because we're lazy
loading members - we add a filter to exclude membership events on the
first call to `CurrentState`.
- Using an optimized query to get the membership events needed to
calculate history visibility
---------
Co-authored-by: kegsay <kegan@matrix.org>
This adds a new admin endpoint `/_dendrite/admin/purgeRoom/{roomID}`. It
completely erases all database entries for a given room ID.
The roomserver will start by clearing all data for that room and then
will generate an output event to notify downstream components (i.e. the
sync API and federation API) to do the same.
It does not currently clear media and it is currently not implemented
for SQLite since it relies on SQL array operations right now.
Co-authored-by: Neil Alexander <neilalexander@users.noreply.github.com>
Co-authored-by: Till Faelligen <2353100+S7evinK@users.noreply.github.com>
The stale device lists table might contain entries for users we don't
share a room with anymore. This now asks the roomserver about left users
and removes those entries from the table.
Co-authored-by: Neil Alexander <neilalexander@users.noreply.github.com>
Makes the following tests pass
```
/upgrade moves remote aliases to the new room
Local and remote users' homeservers remove a room from their public directory on upgrade
```
This optimizes history visibility checks by (mostly) avoiding database
hits.
Possibly solves https://github.com/matrix-org/dendrite/issues/2777
Co-authored-by: Neil Alexander <neilalexander@users.noreply.github.com>
Adds `PUT
/_matrix/client/v3/directory/list/appservice/{networkId}/{roomId}` and
`DELTE
/_matrix/client/v3/directory/list/appservice/{networkId}/{roomId}`
support, as well as the ability to filter `/publicRooms` on networkID
and including all networks.
* Reprocess outliers that were previously rejected
* Might as well do all events this way
* More useful errors
* Fix queries
* Tweak condition
* Don't wrap errors
* Report more useful error
* Flatten error on `r.Queryer.QueryStateAfterEvents`
* Some more debug logging
* Flatten error in `QueryRestrictedJoinAllowed`
* Revert "Flatten error in `QueryRestrictedJoinAllowed`"
This reverts commit 1238b4184c.
* Tweak `QueryStateAfterEvents`
* Handle MissingStateError too
* Scope to room
* Clean up
* Fix the error
* Only apply rejection check to outliers
* Try optimising checking if server is allowed to see event
* Fix error
* Handle case where snapshot NID is 0
* Fix query
* Update SQL
* Clean up `CheckServerAllowedToSeeEvent`
* Not supported on SQLite
* Maybe placate the unit tests
* Review comments
* Try Ristretto cache
* Tweak
* It's beautiful
* Update GMSL
* More strict keyable interface
* Fix that some more
* Make less panicky
* Don't enforce mutability checks for now
* Determine mutability using deep equality
* Tweaks
* Namespace keys
* Make federation caches mutable
* Update cost estimation, add metric
* Update GMSL
* Estimate cost for metrics better
* Reduce counters a bit
* Try caching events
* Some guards
* Try again
* Try this
* Use separate caches for hopefully better hash distribution
* Fix bug with admitting events into cache
* Try to fix bugs
* Check nil
* Try that again
* Preserve order jeezo this is messy
* thanks VS Code for doing exactly the wrong thing
* Try this again
* Be more specific
* aaaaargh
* One more time
* That might be better
* Stronger sorting
* Cache expiries, async publishing of EDUs
* Put it back
* Use a shared cache again
* Cost estimation fixes
* Update ristretto
* Reduce counters a bit
* Clean up a bit
* Update GMSL
* 1GB
* Configurable cache sizees
* Tweaks
* Add `config.DataUnit` for specifying friendly cache sizes
* Various tweaks
* Update GMSL
* Add back some lazy loading caching
* Include key in cost
* Include key in cost
* Tweak max age handling, config key name
* Only register prometheus metrics if requested
* Review comments @S7evinK
* Don't return errors when creating caches (it is better just to crash since otherwise we'll `nil`-pointer exception everywhere)
* Review comments
* Update sample configs
* Update GHA Workflow
* Update Complement images to Go 1.18
* Remove the cache test from the federation API as we no longer guarantee immediate cache admission
* Don't check the caches in the renewal test
* Possibly fix the upgrade tests
* Update to matrix-org/gomatrixserverlib#322
* Update documentation to refer to Go 1.18
