Commit Graph

577 Commits

Author SHA1 Message Date
Erik Johnston
fc2c245a1f
Merge pull request #4666 from matrix-org/erikj/register_login_split
Split out registration to worker
2019-02-18 17:18:06 +00:00
Erik Johnston
32e54b472a Fix kicking guest users in worker mode
When guest_access changes from allowed to forbidden all local guest
users should be kicked from the room. This did not happen when
revocation was received from federation on a worker.

Presumably broken in #4141
2019-02-18 14:16:07 +00:00
Erik Johnston
eb2b8523ae Split out registration to worker
This allows registration to be handled by a worker, though the actual
write to the database still happens on master.

Note: due to the in-memory session map all registration requests must be
handled by the same worker.
2019-02-18 12:12:57 +00:00
Richard van der Hoff
f311018823
Fix errors in acme provisioning (#4648)
* Better logging for errors on startup

* Fix "TypeError: '>' not supported" when starting without an existing
  certificate
* Fix a bug where an existing certificate would be reprovisoned every day
2019-02-14 17:10:36 +00:00
Erik Johnston
309f3bb322
Update synapse/app/_base.py
Co-Authored-By: richvdh <1389908+richvdh@users.noreply.github.com>
2019-02-13 13:24:27 +00:00
Richard van der Hoff
767686af48 Use listen_tcp for the replication listener
Fixes the "can't listen on 0.0.0.0" error. Also makes it more consistent with
what we do elsewhere.
2019-02-13 11:59:04 +00:00
Richard van der Hoff
2a5a15aff8 Improve logging around listening services
I wanted to bring listen_tcp into line with listen_ssl in terms of returning a
list of ports, and wanted to check that was a safe thing to do - hence the
logging in `refresh_certificate`.

Also, pull the 'Synapse now listening' message up to homeserver.py, because it
was being duplicated everywhere else.
2019-02-13 11:58:54 +00:00
Richard van der Hoff
e3a0300431 Special-case the default bind_addresses for metrics listener
turns out it doesn't really support ipv6, so let's hack around that by only
listening on ipv4 by default.
2019-02-13 11:48:56 +00:00
Richard van der Hoff
32b781bfe2
Fix error when loading cert if tls is disabled (#4618)
If TLS is disabled, it should not be an error if no cert is given.

Fixes #4554.
2019-02-12 10:51:31 +00:00
Richard van der Hoff
4fddf8fc77 Infer no_tls from presence of TLS listeners
Rather than have to specify `no_tls` explicitly, infer whether we need to load
the TLS keys etc from whether we have any TLS-enabled listeners.
2019-02-11 21:39:14 +00:00
Richard van der Hoff
9645728619 Don't create server contexts when TLS is disabled
we aren't going to use them anyway.
2019-02-11 21:32:01 +00:00
Richard van der Hoff
086f6f27d4 Logging improvements around TLS certs
Log which file we're reading keys and certs from, and refactor the code a bit
in preparation for other work
2019-02-11 21:02:06 +00:00
Richard van der Hoff
5d27730a73
Move ClientTLSOptionsFactory init out of refresh_certificates (#4611)
It's nothing to do with refreshing the certificates. No idea why it was here.
2019-02-11 18:03:30 +00:00
Amber Brown
6e2a5aa050 ACME Reprovisioning (#4522) 2019-02-11 10:36:26 +00:00
Erik Johnston
b201149c7e
Merge pull request #4420 from matrix-org/jaywink/openid-listener
New listener resource for the federation API "openid/userinfo" endpoint
2019-02-11 09:44:00 +00:00
Amber Brown
9cd33d2f4b
Deduplicate some code in synapse.app (#4567) 2019-02-08 17:25:57 +00:00
Richard van der Hoff
7615a8ced1 ACME config cleanups (#4525)
* Handle listening for ACME requests on IPv6 addresses

the weird url-but-not-actually-a-url-string doesn't handle IPv6 addresses
without extra quoting. Building a string which you are about to parse again
seems like a weird choice. Let's just use listenTCP, which is consistent with
what we do elsewhere.

* Clean up the default ACME config

make it look a bit more consistent with everything else, and tweak the defaults
to listen on port 80.

* newsfile
2019-01-30 14:17:55 +00:00
Erik Johnston
e87d7a4b0f Raise ConfigError instead 2019-01-30 12:48:09 +00:00
Erik Johnston
270f212a2a _listener_http should return a list 2019-01-30 12:14:50 +00:00
Amber Brown
f6813919e8
SIGHUP for TLS cert reloading (#4495) 2019-01-30 11:00:02 +00:00
Amber Brown
5d976c0c7c
Fix worker TLS (#4492)
* load cert

* changelog

* fix
2019-01-28 17:18:33 +00:00
Jason Robinson
6f680241bd Fix flake8 issues
Signed-off-by: Jason Robinson <jasonr@matrix.org>
2019-01-23 10:53:48 +02:00
Amber Brown
6129e52f43
Support ACME for certificate provisioning (#4384) 2019-01-23 19:39:06 +11:00
Jason Robinson
82e13662c0 Split federation OpenID userinfo endpoint out of the federation resource
This allows the OpenID userinfo endpoint to be active even if the
federation resource is not active. The OpenID userinfo endpoint
is called by integration managers to verify user actions using the
client API OpenID access token. Without this verification, the
integration manager cannot know that the access token is valid.

The OpenID userinfo endpoint will be loaded in the case that either
"federation" or "openid" resource is defined. The new "openid"
resource is defaulted to active in default configuration.

Signed-off-by: Jason Robinson <jasonr@matrix.org>
2019-01-23 10:32:41 +02:00
Jason Robinson
a17bac171f Make SynapseHomeServer _http_listener use self.get_reactor()
For all the homeserver classes, only the FrontendProxyServer passes
its reactor when doing the http listen. Looking at previous PR's looks
like this was introduced to make it possible to write a test, otherwise
when you try to run a test with the test homeserver it tries to
do a real bind to a port. Passing the reactor that the homeserver
is instantiated with should probably be the right thing to do anyway?

Signed-off-by: Jason Robinson <jasonr@matrix.org>
2019-01-23 10:32:41 +02:00
Jason Robinson
4f8f41c824 Make FederationReaderServer _http_listen use self.get_reactor()
For all the homeserver classes, only the FrontendProxyServer passes
its reactor when doing the http listen. Looking at previous PR's looks
like this was introduced to make it possible to write a test, otherwise
when you try to run a test with the test homeserver it tries to
do a real bind to a port. Passing the reactor that the homeserver
is instantiated with should probably be the right thing to do anyway?

Signed-off-by: Jason Robinson <jasonr@matrix.org>
2019-01-23 10:32:41 +02:00
Richard van der Hoff
b7c0218812
Check jinja version for consent resource (#4327)
* Raise a ConfigError if an invalid resource is specified

* Require Jinja 2.9 for the consent resource

* changelog
2019-01-07 10:14:31 +00:00
Amber Brown
c26f49a664
Make the dependencies more like a standard Python project and hook up the optional dependencies to setuptools (#4298) 2018-12-22 01:37:26 +11:00
Neil Johnson
7e22cd90f5
ensure can report mau stats when hs.config.mau_stats_only is set (#4305)
* ensure can report mau stats when hs.config.mau_stats_only is set
2018-12-18 14:36:11 +00:00
Richard van der Hoff
f208f608cb Merge branch 'release-v0.34.0' into develop 2018-12-11 15:43:20 +00:00
Richard van der Hoff
188945713e
Merge pull request #4290 from matrix-org/rav/remove_webclient
Stop installing Matrix Console by default
2018-12-11 16:24:15 +01:00
Richard van der Hoff
f537432ef9 Add a welcome page to the static resources
This is largely a precursor for the removal of the bundled webclient. The idea
is to present a page at / which reassures people that something is working, and
to give them some links for next steps.

The welcome page lives at `/_matrix/static/`, so is enabled alongside the other
`static` resources (which, in practice, means the client API is enabled). We'll
redirect to it from `/` if we have nothing better to display there.

It would be nice to have a way to disable it (in the same way that you might
disable the nginx welcome page), but I can't really think of a good way to do
that without a load of ickiness.

It's based on the work done by @krombel for #2601.
2018-12-11 13:26:22 +00:00
Richard van der Hoff
df96177ca7 Stop installing Matrix Console by default
This is based on the work done by @krombel in #2601.
2018-12-11 13:20:33 +00:00
Richard van der Hoff
c7401a697f
Implement SAML2 authentication (#4267)
This implements both a SAML2 metadata endpoint (at
`/_matrix/saml2/metadata.xml`), and a SAML2 response receiver (at
`/_matrix/saml2/authn_response`). If the SAML2 response matches what's been
configured, we complete the SSO login flow by redirecting to the client url
(aka `RelayState` in SAML2 jargon) with a login token.

What we don't yet have is anything to build a SAML2 request and redirect the
user to the identity provider. That is left as an exercise for the reader.
2018-12-07 13:11:11 +01:00
Richard van der Hoff
e8d98466b0
Implement .well-known handling (#4262)
Sometimes it's useful for synapse to generate its own .well-known file.
2018-12-05 14:38:58 +01:00
Travis Ralston
835779f7fb Add option to track MAU stats (but not limit people) (#3830) 2018-11-15 18:08:27 +00:00
Amber Brown
cb7a6b2379
Fix typing being reset causing infinite syncs (#4127) 2018-11-03 00:19:23 +11:00
Amber Brown
f79f454485
Remove deprecated v1 key exchange endpoint (#4119) 2018-10-31 22:29:02 +11:00
Neil Johnson
07126e43a4 Merge branch 'develop' of github.com:matrix-org/synapse into neilj/fix_mau_init 2018-10-24 16:25:39 +01:00
Richard van der Hoff
e0b9d5f0af
Merge pull request #4075 from matrix-org/rav/fix_pusher_logcontexts
Clean up the way logcontexts and threads work in the pushers
2018-10-24 09:53:57 +01:00
Neil Johnson
6105c6101f fix race condiftion in calling initialise_reserved_users 2018-10-23 15:24:58 +01:00
Richard van der Hoff
e7a16c6210 Remove redundant run_as_background_process() from pusherpool
`on_new_notifications` and `on_new_receipts` in `HttpPusher` and `EmailPusher`
now always return synchronously, so we can remove the `defer.gatherResults` on
their results, and the `run_as_background_process` wrappers can be removed too
because the PusherPool methods will now complete quickly enough.
2018-10-22 16:12:11 +01:00
Richard van der Hoff
3e8b02c939 Rename _refresh_pusher
This is public (or at least, called from outside the class), so ought to have a
better name.
2018-10-22 16:12:11 +01:00
Richard van der Hoff
7aea00069c
Merge pull request #4076 from matrix-org/rav/fix_init_logcontexts
Run MAU queries as background processes
2018-10-22 14:46:59 +01:00
Richard van der Hoff
911db96658
Merge pull request #4073 from matrix-org/rav/require_psutil
Make psutil an explicit dependency
2018-10-22 12:33:21 +01:00
Richard van der Hoff
a6f421e812 Run MAU queries as background processes
Fixes #3820
2018-10-20 02:14:35 +01:00
Amber Brown
e1728dfcbe
Make scripts/ and scripts-dev/ pass pyflakes (and the rest of the codebase on py3) (#4068) 2018-10-20 11:16:55 +11:00
Richard van der Hoff
e5b52d0f94 Make psutil an explicit dependency
As of #4027, we require psutil to be installed, so it should be in our
dependency list. We can also remove some of the conditional import code
introduced by #992.

Fixes #4062.
2018-10-19 21:51:15 +01:00
Ivan Shapovalov
06bc8d2fe5 synapse/app: frontend_proxy.py: actually make workers work on py3 2018-10-14 20:08:39 +03:00
Amber Brown
381d2cfdf0
Make workers work on Py3 (#4027) 2018-10-13 00:14:08 +11:00