Currently we use `JsonEncoder.iterencode` to write JSON responses, which ensures that we don't block the main reactor thread when encoding huge objects. The downside to this is that `iterencode` falls back to using a pure Python encoder that is *much* less efficient and can easily burn a lot of CPU for huge responses. To fix this, while still ensuring we don't block the reactor loop, we encode the JSON on a threadpool using the standard `JsonEncoder.encode` functions, which is backed by a C library.
Doing so, however, requires `respond_with_json` to have access to the reactor, which it previously didn't. There are two ways of doing this:
1. threading through the reactor object, which is a bit fiddly as e.g. `DirectServeJsonResource` doesn't currently take a reactor, but is exposed to modules and so is a PITA to change; or
2. expose the reactor in `SynapseRequest`, which requires updating a bunch of servlet types.
I went with the latter as that is just a mechanical change, and I think makes sense as a request already has a reactor associated with it (via its http channel).
Instead of wrapping the JSON into an object, this creates concrete
instances for Transaction and Edu. This allows for improved type
hints and simplified code.
Mostly this involves decorating a few Deferred declarations with extra type hints. We wrap the types in quotes to avoid runtime errors when running against older versions of Twisted that don't have generics on Deferred.
This PR adds a common configuration section for all modules (see docs). These modules are then loaded at startup by the homeserver. Modules register their hooks and web resources using the new `register_[...]_callbacks` and `register_web_resource` methods of the module API.
This is the first of two PRs which seek to address #8518. This first PR lays the groundwork by extending ResponseCache; a second PR (#10158) will update the SyncHandler to actually use it, and fix the bug.
The idea here is that we allow the callback given to ResponseCache.wrap to decide whether its result should be cached or not. We do that by (optionally) passing a ResponseCacheContext into it, which it can modify.
* Make `invalidate` and `invalidate_many` do the same thing
... so that we can do either over the invalidation replication stream, and also
because they always confused me a bit.
* Kill off `invalidate_many`
* changelog
`keylen` seems to be a thing that is frequently incorrectly set, and we don't really need it.
The only time it was used was to figure out if we had removed a subtree in `del_multi`, which we can do better by changing `TreeCache.pop` to return a different type (`TreeCacheNode`).
Commits should be independently reviewable.
- use a tuple rather than a list for the iterable that is passed into the
wrapped function, for performance
- test that we can pass an iterable and that keys are correctly deduped.
We can get away with just catching UnicodeError here.
⋮
+-- ValueError
| +-- UnicodeError
| +-- UnicodeDecodeError
| +-- UnicodeEncodeError
| +-- UnicodeTranslateError
⋮
https://docs.python.org/3/library/exceptions.html#exception-hierarchy
Signed-off-by: Dan Callahan <danc@element.io>
Functionally identical, but more obviously cryptographically secure.
...Explicit is better than implicit?
Avoids needing to know that SystemRandom() implies a CSPRNG, and
complies with the big scary red box on the documentation for random:
> Warning:
> The pseudo-random generators of this module should not be used for
> security purposes. For security or cryptographic uses, see the
> secrets module.
https://docs.python.org/3/library/random.html
Signed-off-by: Dan Callahan <danc@element.io>
* tests for push rule pattern matching
* tests for acl pattern matching
* factor out common `re.escape`
* Factor out common re.compile
* Factor out common anchoring code
* add word_boundary support to `glob_to_regex`
* Use `glob_to_regex` in push rule evaluator
NB that this drops support for character classes. I don't think anyone ever
used them.
* Improve efficiency of globs with multiple wildcards
The idea here is that we compress multiple `*` globs into a single `.*`. We
also need to consider `?`, since `*?*` is as hard to implement efficiently as
`**`.
* add assertion on regex pattern
* Fix mypy
* Simplify glob_to_regex
* Inline the glob_to_regex helper function
Signed-off-by: Dan Callahan <danc@element.io>
* Moar comments
Signed-off-by: Dan Callahan <danc@element.io>
Co-authored-by: Dan Callahan <danc@element.io>
I went through and removed a bunch of cruft that was lying around for compatibility with old Python versions. This PR also will now prevent Synapse from starting unless you're running Python 3.6+.
As far as I can tell our logging contexts are meant to log the request ID, or sometimes the request ID followed by a suffix (this is generally stored in the name field of LoggingContext). There's also code to log the name@memory location, but I'm not sure this is ever used.
This simplifies the code paths to require every logging context to have a name and use that in logging. For sub-contexts (created via nested_logging_contexts, defer_to_threadpool, Measure) we use the current context's str (which becomes their name or the string "sentinel") and then potentially modify that (e.g. add a suffix).
Part of #9744
Removes all redundant `# -*- coding: utf-8 -*-` lines from files, as python 3 automatically reads source code as utf-8 now.
`Signed-off-by: Jonathan de Jong <jonathan@automatia.nl>`
Running `dmypy run` will do a `mypy` check while spinning up a daemon
that makes rerunning `dmypy run` a lot faster.
`dmypy` doesn't support `follow_imports = silent` and has
`local_partial_types` enabled, so this PR enables those options and
fixes the issues that were newly raised. Note that `local_partial_types`
will be enabled by default in upcoming mypy releases.
This great big stack of commits is a a whole load of hoop-jumping to make it easier to store additional values in login tokens, and then to actually store the SSO Identity Provider in the login token. (Making use of that data will follow in a subsequent PR.)
This fixes#8518 by adding a conditional check on `SyncResult` in a function when `prev_stream_token == current_stream_token`, as a sanity check. In `CachedResponse.set.<remove>()`, the result is immediately popped from the cache if the conditional function returns "false".
This prevents the caching of a timed-out `SyncResult` (that has `next_key` as the stream key that produced that `SyncResult`). The cache is prevented from returning a `SyncResult` that makes the client request the same stream key over and over again, effectively making it stuck in a loop of requesting and getting a response immediately for as long as the cache keeps those values.
Signed-off-by: Jonathan de Jong <jonathan@automatia.nl>
- Update black version to the latest
- Run black auto formatting over the codebase
- Run autoformatting according to [`docs/code_style.md
`](80d6dc9783/docs/code_style.md)
- Update `code_style.md` docs around installing black to use the correct version
==============================
Features
--------
- Further improvements to the user experience of registration via single sign-on. ([\#9297](https://github.com/matrix-org/synapse/issues/9297))
Bugfixes
--------
- Fix ratelimiting introduced in v1.27.0rc1 for invites to respect the `ratelimit` flag on application services. ([\#9302](https://github.com/matrix-org/synapse/issues/9302))
- Do not automatically calculate `public_baseurl` since it can be wrong in some situations. Reverts behaviour introduced in v1.26.0. ([\#9313](https://github.com/matrix-org/synapse/issues/9313))
Improved Documentation
----------------------
- Clarify the sample configuration for changes made to the template loading code. ([\#9310](https://github.com/matrix-org/synapse/issues/9310))
-----BEGIN PGP SIGNATURE-----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=NSyp
-----END PGP SIGNATURE-----
Merge tag 'v1.27.0rc2' into develop
Synapse 1.27.0rc2 (2021-02-11)
==============================
Features
--------
- Further improvements to the user experience of registration via single sign-on. ([\#9297](https://github.com/matrix-org/synapse/issues/9297))
Bugfixes
--------
- Fix ratelimiting introduced in v1.27.0rc1 for invites to respect the `ratelimit` flag on application services. ([\#9302](https://github.com/matrix-org/synapse/issues/9302))
- Do not automatically calculate `public_baseurl` since it can be wrong in some situations. Reverts behaviour introduced in v1.26.0. ([\#9313](https://github.com/matrix-org/synapse/issues/9313))
Improved Documentation
----------------------
- Clarify the sample configuration for changes made to the template loading code. ([\#9310](https://github.com/matrix-org/synapse/issues/9310))
There's some prelimiary work here to pull out the construction of a jinja environment to a separate function.
I wanted to load the template at display time rather than load time, so that it's easy to update on the fly. Honestly, I think we should do this with all our templates: the risk of ending up with malformed templates is far outweighed by the improved turnaround time for an admin trying to update them.
If a Synapse module's config block were empty in YAML, thus being translated to a `Nonetype` in Python, then some modules could fail as that None ends up getting passed to their `parse_config` method. Modules are expected to accept a `dict` instead.
This PR ensures that if the user does end up specifying an empty config block (such as what [the default oidc config in the sample config](5310808d3b/docs/sample_config.yaml (L1816-L1845)) states) then `None` is not passed to the module. An empty dict is passed instead.
This code assumes that no existing modules are relying on receiving a `None` config block, but I'd really hope that they aren't.
`distutils` is pretty much deprecated these days, and replaced with
`setuptools`. It's also annoying because it's you can't `pip install` it, and
it's hard to figure out which debian package we should depend on to make sure
it's there.
Since we only use it for a tiny function anyway, let's just vendor said
function into our codebase.
The idea is that the parse_config method of extension modules can raise either a ConfigError or a JsonValidationError,
and it will be magically turned into a legible error message. There's a few components to it:
* Separating the "path" and the "message" parts of a ConfigError, so that we can fiddle with the path bit to turn it
into an absolute path.
* Generally improving the way ConfigErrors get printed.
* Passing in the config path to load_module so that it can wrap any exceptions that get caught appropriately.
Not being able to serialise `frozendicts` is fragile, and it's annoying to have
to think about which serialiser you want. There's no real downside to
supporting frozendicts, so let's just have one json encoder.
This allows trailing commas in multi-line arg lists.
Minor, but we might as well keep our formatting current with regard to
our minimum supported Python version.
Signed-off-by: Dan Callahan <danc@element.io>
* 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
* Remove `on_timeout_cancel` from `timeout_deferred`
The `on_timeout_cancel` param to `timeout_deferred` wasn't always called on a
timeout (in particular if the canceller raised an exception), so it was
unreliable. It was also only used in one place, and to be honest it's easier to
do what it does a different way.
* Fix handling of connection timeouts in outgoing http requests
Turns out that if we get a timeout during connection, then a different
exception is raised, which wasn't always handled correctly.
To fix it, catch the exception in SimpleHttpClient and turn it into a
RequestTimedOutError (which is already a documented exception).
Also add a description to RequestTimedOutError so that we can see which stage
it failed at.
* Fix incorrect handling of timeouts reading federation responses
This was trapping the wrong sort of TimeoutError, so was never being hit.
The effect was relatively minor, but we should fix this so that it does the
expected thing.
* Fix inconsistent handling of `timeout` param between methods
`get_json`, `put_json` and `delete_json` were applying a different timeout to
the response body to `post_json`; bring them in line and test.
Co-authored-by: Patrick Cloke <clokep@users.noreply.github.com>
Co-authored-by: Erik Johnston <erik@matrix.org>
slots use less memory (and attribute access is faster) while slightly
limiting the flexibility of the class attributes. This focuses on objects
which are instantiated "often" and for short periods of time.
==============================
Bugfixes
--------
- Fix a bug introduced in v1.20.0rc1 where the wrong exception was raised when invalid JSON data is encountered. ([\#8291](https://github.com/matrix-org/synapse/issues/8291))
-----BEGIN PGP SIGNATURE-----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=e2T1
-----END PGP SIGNATURE-----
Merge tag 'v1.20.0rc3' into develop
Synapse 1.20.0rc3 (2020-09-11)
==============================
Bugfixes
--------
- Fix a bug introduced in v1.20.0rc1 where the wrong exception was raised when invalid JSON data is encountered. ([\#8291](https://github.com/matrix-org/synapse/issues/8291))
By importing from canonicaljson the simplejson module was still being used
in some situations. After this change the std lib json is consistenty used
throughout Synapse.
This has long been something I've wanted to do. Basically the `Daemonize` code
is both too flexible and not flexible enough, in that it offers a bunch of
features that we don't use (changing UID, closing FDs in the child, logging to
syslog) and doesn't offer a bunch that we could do with (redirecting stdout/err
to a file instead of /dev/null; having the parent not exit until the child is
running).
As a first step, I've lifted the Daemonize code and removed the bits we don't
use. This should be a non-functional change. Fixing everything else will come
later.
While working on https://github.com/matrix-org/synapse/issues/5665 I found myself digging into the `Ratelimiter` class and seeing that it was both:
* Rather undocumented, and
* causing a *lot* of config checks
This PR attempts to refactor and comment the `Ratelimiter` class, as well as encourage config file accesses to only be done at instantiation.
Best to be reviewed commit-by-commit.
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
Other parts of the code (such as the StreamChangeCache) assume that there will
not be multiple changes with the same stream id.
This code was introduced in #7024, and I hope this fixes#7206.
* 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.
A lot of the things we log at INFO are now a bit superfluous, so lets
make them DEBUG logs to reduce the amount we log by default.
Co-Authored-By: Brendan Abolivier <babolivier@matrix.org>
Co-authored-by: Brendan Abolivier <github@brendanabolivier.com>
`Measure` incorrectly assumed that it was the only thing being done by the parent `LoggingContext`. For instance, during a "renew group attestations" operation, hundreds of `outbound_request` calls could take place in parallel, all using the same `LoggingContext`. This would mean that any resources used during *any* of those calls would be reported against *all* of them, producing wildly inaccurate results.
Instead, we now give each `Measure` block its own `LoggingContext` (using the parent `LoggingContext` mechanism to ensure that the log lines look correct and that the metrics are ultimately propogated to the top level for reporting against requests/backgrond tasks).
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.
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).
Fixes a bug where the default attribute maps were prioritised over
user-specified ones, resulting in incorrect mappings.
The problem is that if you call SPConfig.load() multiple times, it adds new
attribute mappers to a list. So by calling it with the default config first,
and then the user-specified config, we would always get the default mappers
before the user-specified mappers.
To solve this, let's merge the config dicts first, and then pass them to
SPConfig.
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.
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.
The version of a module isn't going to change over the lifetime of the
process (assuming no funky hot reloading is going on, which it isn't),
so let's just cache the result to avoid spawning lots of git
subprocesses.
Fixes#5672.
- 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
Closes#4583
Does slightly less than #5045, which prevented a room from being upgraded multiple times, one after another. This PR still allows that, but just prevents two from happening at the same time.
Mostly just to mitigate the fact that servers are slow and it can take a moment for the room upgrade to actually complete. We don't want people sending another request to upgrade the room when really they just thought the first didn't go through.
Hillery Shay