The idea here is that if an instance persists an event via the replication HTTP API it can return before we receive that event over replication, which can lead to races where code assumes that persisting an event immediately updates various caches (e.g. current state of the room).
Most of Synapse doesn't hit such races, so we don't do the waiting automagically, instead we do so where necessary to avoid unnecessary delays. We may decide to change our minds here if it turns out there are a lot of subtle races going on.
People probably want to look at this commit by commit.
When a call to `user_device_resync` fails, we don't currently mark the remote user's device list as out of sync, nor do we retry to sync it.
https://github.com/matrix-org/synapse/pull/6776 introduced some code infrastructure to mark device lists as stale/out of sync.
This commit uses that code infrastructure to mark device lists as out of sync if processing an incoming device list update makes the device handler realise that the device list is out of sync, but we can't resync right now.
It also adds a looping call to retry all failed resync every 30s. This shouldn't cause too much spam in the logs as this commit also removes the "Failed to handle device list update for..." warning logs when catching `NotRetryingDestination`.
Fixes#7418
Make sure that the AccountDataStream presents complete updates, in the right
order.
This is much the same fix as #7337 and #7358, but applied to a different stream.
==============================
Bugfixes
--------
- Fix a long-standing bug which could cause messages not to be sent over federation, when state events with state keys matching user IDs (such as custom user statuses) were received. ([\#7376](https://github.com/matrix-org/synapse/issues/7376))
- Restore compatibility with non-compliant clients during the user interactive authentication process, fixing a problem introduced in v1.13.0rc1. ([\#7483](https://github.com/matrix-org/synapse/issues/7483))
Internal Changes
----------------
- Fix linting errors in new version of Flake8. ([\#7470](https://github.com/matrix-org/synapse/issues/7470))
-----BEGIN PGP SIGNATURE-----
iQEzBAABCAAdFiEEv27Axt/F4vrTL/8QOSor00I9eP8FAl69IQ8ACgkQOSor00I9
eP87lAf8DK+v6cs2U0BoD5opzQ7ZazJT6JYTmnMBaTzHU6Wx20V2ttkF7Vpwm3WU
Zsz0048tdYtHFyYBQ1kF5RNIBBJwV8SA/QUcPkR7FVpwZMLR2q4aJn0EE7kC9OMf
tYsmdbHeBdyfLXpXzazxWlgHquLyEIt52ykAcCphjx/Jl2fAExFEhtfsxpECoJ2f
8Dqhjg3WFjd6QWU6AFkElbwHUYCdIWdJOcsC8N1p8OvBmDz5QXv/RlYipHE00Cpx
QQQOgEjdRc6dlz2mbetMklnfII3p2kO9bzNdmEpOzT0Zt7nFaGdntW4I1QA0yJfa
gows9bYMzhqYk7YSiyTYOZ4qyavVtw==
=N/zZ
-----END PGP SIGNATURE-----
Merge tag 'v1.13.0rc2' into develop
Synapse 1.13.0rc2 (2020-05-14)
==============================
Bugfixes
--------
- Fix a long-standing bug which could cause messages not to be sent over federation, when state events with state keys matching user IDs (such as custom user statuses) were received. ([\#7376](https://github.com/matrix-org/synapse/issues/7376))
- Restore compatibility with non-compliant clients during the user interactive authentication process, fixing a problem introduced in v1.13.0rc1. ([\#7483](https://github.com/matrix-org/synapse/issues/7483))
Internal Changes
----------------
- Fix linting errors in new version of Flake8. ([\#7470](https://github.com/matrix-org/synapse/issues/7470))
Fix a bug where the `get_joined_users` cache could be corrupted by custom
status events (or other state events with a state_key matching the user ID).
The bug was introduced by #2229, but has largely gone unnoticed since then.
Fixes#7099, #7373.
The aim here is to get to a stage where we have a `PersistEventStore` that holds all the write methods used during event persistence, so that we can take that class out of the `DataStore` mixin and instansiate it separately. This will allow us to instansiate it on processes other than master, while also ensuring it is only available on processes that are configured to write to events stream.
This is a bit of an architectural change, where we end up with multiple classes per data store (rather than one per data store we have now). We end up having:
1. Storage classes that provide high level APIs that can talk to multiple data stores.
2. Data store modules that consist of classes that must point at the same database instance.
3. Classes in a data store that can be instantiated on processes depending on config.
* release-v1.13.0:
Don't UPGRADE database rows
RST indenting
Put rollback instructions in upgrade notes
Fix changelog typo
Oh yeah, RST
Absolute URL it is then
Fix upgrade notes link
Provide summary of upgrade issues in changelog. Fix )
Move next version notes from changelog to upgrade notes
Changelog fixes
1.13.0rc1
Documentation on setting up redis (#7446)
Rework UI Auth session validation for registration (#7455)
Fix errors from malformed log line (#7454)
Drop support for redis.dbid (#7450)
For in memory streams when fetching updates on workers we need to query the source of the stream, which currently is hard coded to be master. This PR threads through the source instance we received via `POSITION` through to the update function in each stream, which can then be passed to the replication client for in memory streams.
We move the processing of typing and federation replication traffic into their handlers so that `Stream.current_token()` points to a valid token. This allows us to remove `get_streams_to_replicate()` and `stream_positions()`.
By persisting the user interactive authentication sessions to the database, this fixes
situations where a user hits different works throughout their auth session and also
allows sessions to persist through restarts of Synapse.
This is primarily for allowing us to send those commands from workers, but for now simply allows us to ignore echoed RDATA/POSITION commands that we sent (we get echoes of sent commands when using redis). Currently we log a WARNING on the master process every time we receive an echoed RDATA.
For direct TCP connections we need the master to relay REMOTE_SERVER_UP
commands to the other connections so that all instances get notified
about it. The old implementation just relayed to all connections,
assuming that sending back to the original sender of the command was
safe. This is not true for redis, where commands sent get echoed back to
the sender, which was causing master to effectively infinite loop
sending and then re-receiving REMOTE_SERVER_UP commands that it sent.
The fix is to ensure that we only relay to *other* connections and not
to the connection we received the notification from.
Fixes#7334.
* Factor out functions for injecting events into database
I want to add some more flexibility to the tools for injecting events into the
database, and I don't want to clutter up HomeserverTestCase with them, so let's
factor them out to a new file.
* Rework TestReplicationDataHandler
This wasn't very easy to work with: the mock wrapping was largely superfluous,
and it's useful to be able to inspect the received rows, and clear out the
received list.
* Fix AssertionErrors being thrown by EventsStream
Part of the problem was that there was an off-by-one error in the assertion,
but also the limit logic was too simple. Fix it all up and add some tests.
Specifically some tests for the typing stream, which means we test streams that fetch missing updates via HTTP (rather than via the DB).
We also shuffle things around a bit so that we create two separate `HomeServer` objects, rather than trying to insert a slaved store into places.
Note: `test_typing.py` is heavily inspired by `test_receipts.py`
First some background: StreamChangeCache is used to keep track of what "entities" have
changed since a given stream ID. So for example, we might use it to keep track of when the last
to-device message for a given user was received [1], and hence whether we need to pull any to-device messages from the database on a sync [2].
Now, it turns out that StreamChangeCache didn't support more than one thing being changed at
a given stream_id (this was part of the problem with #7206). However, it's entirely valid to send
to-device messages to more than one user at a time.
As it turns out, this did in fact work, because *some* methods of StreamChangeCache coped
ok with having multiple things changing on the same stream ID, and it seems we never actually
use the methods which don't work on the stream change caches where we allow multiple
changes at the same stream ID. But that feels horribly fragile, hence: let's update
StreamChangeCache to properly support this, and add some typing and some more tests while
we're at it.
[1]: https://github.com/matrix-org/synapse/blob/release-v1.12.3/synapse/storage/data_stores/main/deviceinbox.py#L301
[2]: https://github.com/matrix-org/synapse/blob/release-v1.12.3/synapse/storage/data_stores/main/deviceinbox.py#L47-L51
The aim here is to move the command handling out of the TCP protocol classes and to also merge the client and server command handling (so that we can reuse them for redis protocol). This PR simply moves the client paths to the new `ReplicationCommandHandler`, a future PR will move the server paths too.
Fixes#6815
Before figuring out whether we should alert a user on MAU, we call get_notice_room_for_user to get some info on the existing server notices room for this user. This function, if the room doesn't exist, creates it and invites the user in it. This means that, if we decide later that no server notice is needed, the user gets invited in a room with no message in it. This happens at every restart of the server, since the room ID returned by get_notice_room_for_user is cached.
This PR fixes that by moving the inviting bit to a dedicated function, that's only called when the server actually needs to send a notice to the user. A potential issue with this approach is that the room that's created by get_notice_room_for_user doesn't match how that same function looks for an existing room (i.e. it creates a room that doesn't have an invite or a join for the current user in it, so it could lead to a new room being created each time a user syncs), but I'm not sure this is a problem given it's cached until the server restarts, so that function won't run very often.
It also renames get_notice_room_for_user into get_or_create_notice_room_for_user to make what it does clearer.
Occasionally we could get a federation device list update transaction which
looked like:
```
[
{'edu_type': 'm.device_list_update', 'content': {'user_id': '@user:test', 'device_id': 'D2', 'prev_id': [], 'stream_id': 12, 'deleted': True}},
{'edu_type': 'm.device_list_update', 'content': {'user_id': '@user:test', 'device_id': 'D1', 'prev_id': [12], 'stream_id': 11, 'deleted': True}},
{'edu_type': 'm.device_list_update', 'content': {'user_id': '@user:test', 'device_id': 'D3', 'prev_id': [11], 'stream_id': 13, 'deleted': True}}
]
```
Having `stream_ids` which are lower than `prev_ids` looks odd. It might work
(I'm not actually sure), but in any case it doesn't seem like a reasonable
thing to expect other implementations to support.
That fallback sets the redirect URL to itself (so it can process the login
token then return gracefully to the client). This would make it pointless to
ask the user for confirmation, since the URL the confirmation page would be
showing wouldn't be the client's.
This changes the replication protocol so that the server does not send down `RDATA` for rows that happened before the client connected. Instead, the server will send a `POSITION` and clients then query the database (or master out of band) to get up to date.
* Pull Sentinel out of LoggingContext
... and drop a few unnecessary references to it
* Factor out LoggingContext.current_context
move `current_context` and `set_context` out to top-level functions.
Mostly this means that I can more easily trace what's actually referring to
LoggingContext, but I think it's generally neater.
* move copy-to-parent into `stop`
this really just makes `start` and `stop` more symetric. It also means that it
behaves correctly if you manually `set_log_context` rather than using the
context manager.
* Replace `LoggingContext.alive` with `finished`
Turn `alive` into `finished` and make it a bit better defined.
* Add 'device_lists_outbound_pokes' as extra table.
This makes sure we check all the relevant tables to get the current max
stream ID.
Currently not doing so isn't problematic as the max stream ID in
`device_lists_outbound_pokes` is the same as in `device_lists_stream`,
however that will change.
* Change device lists stream to have one row per id.
This will make it possible to process the streams more incrementally,
avoiding having to process large chunks at once.
* Change device list replication to match new semantics.
Instead of sending down batches of user ID/host tuples, send down a row
per entity (user ID or host).
* Newsfile
* Remove handling of multiple rows per ID
* Fix worker handling
* Comments from review
It was originally implemented by pulling the full auth chain of all
state sets out of the database and doing set comparison. However, that
can take a lot work if the state and auth chains are large.
Instead, lets try and fetch the auth chains at the same time and
calculate the difference on the fly, allowing us to bail early if all
the auth chains converge. Assuming that the auth chains do converge more
often than not, this should improve performance. Hopefully.
When we get an invite over federation, store the room version in the rooms table.
The general idea here is that, when we pull the invite out again, we'll want to know what room_version it belongs to (so that we can later redact it if need be). So we need to store it somewhere...
The state res v2 algorithm only cares about the difference between auth
chains, so we can pass in the known common state to the `get_auth_chain`
storage function so that it can ignore those events.
* Reject device display names that are too long.
Too long is currently defined as 100 characters in length.
* Add a regression test for rejecting a too long device display name.
These are easier to work with than the strings and we normally have one around.
This fixes `FederationHander._persist_auth_tree` which was passing a
RoomVersion object into event_auth.check instead of a string.
There are quite a few places that we assume that a redaction event has a
corresponding `redacts` key, which is not always the case. So lets
cheekily make it so that event.redacts just returns None instead.
* Port synapse.replication.tcp to async/await
* Newsfile
* Correctly document type of on_<FOO> functions as async
* Don't be overenthusiastic with the asyncing....
Currently we rely on `current_state_events` to figure out what rooms a
user was in and their last membership event in there. However, if the
server leaves the room then the table may be cleaned up and that
information is lost. So lets add a table that separately holds that
information.
This was ill-advised. We can't modify verify_keys here, because the response
object has already been signed by the requested key.
Furthermore, it's somewhat unnecessary because existing versions of Synapse
(which get upset that the notary key isn't present in verify_keys) will fall
back to a direct fetch via `/key/v2/server`.
Also: more tests for fetching keys via perspectives: it would be nice if we actually tested when our fetcher can't talk to our notary impl.
* Kill off redundant SynapseRequestFactory
We already get the Site via the Channel, so there's no need for a dedicated
RequestFactory: we can just use the right constructor.
* Workaround for error when fetching notary's own key
As a notary server, when we return our own keys, include all of our signing
keys in verify_keys.
This is a workaround for #6596.
* Remove redundant python2 support code
`str.decode()` doesn't exist on python3, so presumably this code was doing
nothing
* Filter out pushers with corrupt data
When we get a row with unparsable json, drop the row, rather than returning a
row with null `data`, which will then cause an explosion later on.
* Improve logging when we can't start a pusher
Log the ID to help us understand the problem
* Make email pusher setup more robust
We know we'll have a `data` member, since that comes from the database. What we
*don't* know is if that is a dict, and if that has a `brand` member, and if
that member is a string.
Implement part [MSC2228](https://github.com/matrix-org/matrix-doc/pull/2228). The parts that differ are:
* the feature is hidden behind a configuration flag (`enable_ephemeral_messages`)
* self-destruction doesn't happen for state events
* only implement support for the `m.self_destruct_after` field (not the `m.self_destruct` one)
* doesn't send synthetic redactions to clients because for this specific case we consider the clients to be able to destroy an event themselves, instead we just censor it (by pruning its JSON) in the database
Fixes a bug where rejected events were persisted with the wrong state group.
Also fixes an occasional internal-server-error when receiving events over
federation which are rejected and (possibly because they are
backwards-extremities) have no prev_group.
Fixes#6289.
The intention here is to make it clearer which fields we can expect to be
populated when: notably, that the _event_type etc aren't used for the
synchronous impl of EventContext.
The `http_proxy` and `HTTPS_PROXY` env vars can be set to a `host[:port]` value which should point to a proxy.
The address of the proxy should be excluded from IP blacklists such as the `url_preview_ip_range_blacklist`.
The proxy will then be used for
* push
* url previews
* phone-home stats
* recaptcha validation
* CAS auth validation
It will *not* be used for:
* Application Services
* Identity servers
* Outbound federation
* In worker configurations, connections from workers to masters
Fixes#4198.
This makes it easier to use in an async/await world.
Also fixes a bug where cache descriptors would occaisonally return a raw
value rather than a deferred.
* Fix presence timeouts when synchrotron restarts.
Handling timeouts would fail if there was an external process that had
timed out, e.g. a synchrotron restarting. This was due to a couple of
variable name typoes.
Fixes#3715.
This fixed the weirdness of 400 vs 404 as http status code in the case
the filter id is not known by the server.
As e.g. matrix-js-sdk expects 404 to catch this situation this leads
to unwanted behaviour.
Hopefully this will fix the occasional failures we were seeing in the room directory.
The problem was that events are not necessarily persisted (and `current_state_delta_stream` updated) in the same order as their stream_id. So for instance current_state_delta 9 might be persisted *before* current_state_delta 8. Then, when the room stats saw stream_id 9, it assumed it had done everything up to 9, and never came back to do stream_id 8.
We can solve this easily by only processing up to the stream_id where we know all events have been persisted.
We have set the max retry interval to a value larger than a postgres or
sqlite int can hold, which caused exceptions when updating the
destinations table.
To fix postgres we need to change the column to a bigint, and for sqlite
we lower the max interval to 2**62 (which is still incredibly long).
Pull the checkers out to their own classes, rather than having them lost in a
massive 1000-line class which does everything.
This is also preparation for some more intelligent advertising of flows, as per #6100
because, frankly, it looked like it was written by an axe-murderer.
This should be a non-functional change, except that where `m.login.dummy` was
previously advertised *before* `m.login.terms`, it will now be advertised
afterwards. AFAICT that should have no effect, and will be more consistent with
the flows that involve passing a 3pid.
This is a) simpler than querying user_ips directly and b) means we can
purge older entries from user_ips without losing the required info.
The storage functions now no longer return the access_token, since it
was unused.
This allows support users to be created even on MAU limits via
the admin API. Support users are excluded from MAU after creation,
so it makes sense to exclude them in creation - except if the
whole host is in disabled state.
Signed-off-by: Jason Robinson <jasonr@matrix.org>
Previously if the first registered user was a "support" or "bot" user,
when the first real user registers, the auto-join rooms were not
created.
Fix to exclude non-real (ie users with a special user type) users
when counting how many users there are to determine whether we should
auto-create a room.
Signed-off-by: Jason Robinson <jasonr@matrix.org>
Python will return a tuple whether there are parentheses around the returned values or not.
I'm just sick of my editor complaining about this all over the place :)
Template config files
* Imagine a system composed entirely of x, y, z etc and the basic operations..
Wait George, why XOR? Why not just neq?
George: Eh, I didn't think of that..
Co-Authored-By: Erik Johnston <erik@matrix.org>
Get rid of the labyrinthine `recoverer_fn` code, and clean up the startup code
(it seemed to be previously inexplicably split between
`ApplicationServiceScheduler.start` and `_Recoverer.start`).
Add some docstrings too.
This refactors MatrixFederationAgent to move the SRV lookup into the
endpoint code, this has two benefits:
1. Its easier to retry different host/ports in the same way as
HostnameEndpoint.
2. We avoid SRV lookups if we have a free connection in the pool
If we have recently seen a valid well-known for a domain we want to
retry on (non-final) errors a few times, to handle temporary blips in
networking/etc.
This gives a bit of a grace period where we can attempt to refetch a
remote `well-known`, while still using the cached result if that fails.
Hopefully this will make the well-known resolution a bit more torelant
of failures, rather than it immediately treating failures as "no result"
and caching that for an hour.
It costs both us and the remote server for us to fetch the well known
for every single request we send, so we add a minimum cache period. This
is set to 5m so that we still honour the basic premise of "refetch
frequently".
The `expire_access_token` didn't do what it sounded like it should do. What it
actually did was make Synapse enforce the 'time' caveat on macaroons used as
access tokens, but since our access token macaroons never contained such a
caveat, it was always a no-op.
(The code to add 'time' caveats was removed back in v0.18.5, in #1656)
There was some inconsistent behaviour in the caching layer around how
exceptions were handled - particularly synchronously-thrown ones.
This seems to be most easily handled by pushing the creation of
ObservableDeferreds down from CacheDescriptor to the Cache.
--------
- Fix a regression introduced in v1.2.0rc1 which led to incorrect labels on some prometheus metrics. ([\#5734](https://github.com/matrix-org/synapse/issues/5734))
-----BEGIN PGP SIGNATURE-----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=+JBn
-----END PGP SIGNATURE-----
Merge tag 'v1.2.0rc2' into develop
Bugfixes
--------
- Fix a regression introduced in v1.2.0rc1 which led to incorrect labels on some prometheus metrics. ([\#5734](https://github.com/matrix-org/synapse/issues/5734))
* Fix servlet metric names
Co-Authored-By: Richard van der Hoff <1389908+richvdh@users.noreply.github.com>
* Remove redundant check
* Cover all return paths
We can now use `_get_events_from_cache_or_db` rather than going right back to
the database, which means that (a) we can benefit from caching, and (b) it
opens the way forward to more extensive checks on the original event.
We now always require the original event to exist before we will serve up a
redaction.
First of all, let's get rid of `TOKEN_NOT_FOUND_HTTP_STATUS`. It was a hack we
did at one point when it was possible to return either a 403 or a 401 if the
creds were missing. We always return a 401 in these cases now (thankfully), so
it's not needed.
Let's also stop abusing `AuthError` for these cases. Honestly they have nothing
that relates them to the other places that `AuthError` is used, other than the
fact that they are loosely under the 'Auth' banner. It makes no sense for them
to share exception classes.
Instead, let's add a couple of new exception classes: `InvalidClientTokenError`
and `MissingClientTokenError`, for the `M_UNKNOWN_TOKEN` and `M_MISSING_TOKEN`
cases respectively - and an `InvalidClientCredentialsError` base class for the
two of them.
this is only used in one place, so it's clearer if we inline it and reduce the
API surface.
Also, fixes a buglet where we would create an access token even if we were
about to block the user (we would never return the AT, so the user could never
use it, but it was still created and added to the db.)
When asking for the relations of an event, include the original event in the response. This will mostly be used for efficiently showing edit history, but could be useful in other circumstances.
Nothing uses this now, so we can remove the dead code, and clean up the
API.
Since we're changing the shape of the return value anyway, we take the
opportunity to give the method a better name.
- Put the default window_size back to 1000ms (broken by #5181)
- Make the `rc_federation` config actually do something
- fix an off-by-one error in the 'concurrent' limit
- Avoid creating an unused `_PerHostRatelimiter` object for every single
incoming request
Adds new config option `cleanup_extremities_with_dummy_events` which
periodically sends dummy events to rooms with more than 10 extremities.
THIS IS REALLY EXPERIMENTAL.
Some keys are stored in the synapse database with a null valid_until_ms
which caused an exception to be thrown when using that key. We fix this
by treating nulls as zeroes, i.e. they keys will match verification
requests with a minimum_valid_until_ms of zero (i.e. don't validate ts)
but will not match requests with a non-zero minimum_valid_until_ms.
Fixes#5391.
Sends password reset emails from the homeserver instead of proxying to the identity server. This is now the default behaviour for security reasons. If you wish to continue proxying password reset requests to the identity server you must now enable the email.trust_identity_server_for_password_resets option.
This PR is a culmination of 3 smaller PRs which have each been separately reviewed:
* #5308
* #5345
* #5368
There are a few changes going on here:
* We make checking the signature on a key server response optional: if no
verify_keys are specified, we trust to TLS to validate the connection.
* We change the default config so that it does not require responses to be
signed by the old key.
* We replace the old 'perspectives' config with 'trusted_key_servers', which
is also formatted slightly differently.
* We emit a warning to the logs every time we trust a key server response
signed by the old key.
* Fix background updates to handle redactions/rejections
In background updates based on current state delta stream we need to
handle that we may not have all the events (or at least that
`get_events` may raise an exception).
Also:
* rename VerifyKeyRequest->VerifyJsonRequest
* calculate key_ids on VerifyJsonRequest construction
* refactor things to pass around VerifyJsonRequests instead of 4-tuples
When handling incoming federation requests, make sure that we have an
up-to-date copy of the signing key.
We do not yet enforce the validity period for event signatures.
When enabling the account validity feature, Synapse will look at startup for registered account without an expiration date, and will set one equals to 'now + validity_period' for them. On large servers, it can mean that a large number of users will have the same expiration date, which means that they will all be sent a renewal email at the same time, which isn't ideal.
In order to mitigate this, this PR allows server admins to define a 'max_delta' so that the expiration date is a random value in the [now + validity_period ; now + validity_period + max_delta] range. This allows renewal emails to be progressively sent over a configured period instead of being sent all in one big batch.
The list of server names was redundant, since it was equivalent to the keys on
the server_to_deferred map. This reduces the number of large lists being passed
around, and has the benefit of deduplicating the entries in `wait_on`.
