diff options
| -rw-r--r-- | INSTALL.md | 42 | ||||
| -rw-r--r-- | README.rst | 100 | ||||
| -rw-r--r-- | changelog.d/4614.feature | 1 | ||||
| -rw-r--r-- | changelog.d/4618.bugfix | 1 | ||||
| -rw-r--r-- | changelog.d/4621.misc | 1 | ||||
| -rw-r--r-- | changelog.d/4625.bugfix | 1 | ||||
| -rw-r--r-- | changelog.d/4626.misc | 1 | ||||
| -rw-r--r-- | synapse/app/_base.py | 5 | ||||
| -rw-r--r-- | synapse/config/_base.py | 9 | ||||
| -rw-r--r-- | synapse/config/server.py | 48 | ||||
| -rw-r--r-- | synapse/config/tls.py | 63 | ||||
| -rw-r--r-- | synapse/storage/client_ips.py | 63 | ||||
| -rw-r--r-- | tests/config/test_tls.py | 2 |
13 files changed, 219 insertions, 118 deletions
diff --git a/INSTALL.md b/INSTALL.md index e496a13b21..fb6a5e4e99 100644 --- a/INSTALL.md +++ b/INSTALL.md @@ -350,18 +350,34 @@ Once you have installed synapse as above, you will need to configure it. ## TLS certificates -The default configuration exposes two HTTP ports: 8008 and 8448. Port 8008 is -configured without TLS; it should be behind a reverse proxy for TLS/SSL -termination on port 443 which in turn should be used for clients. Port 8448 -is configured to use TLS for Federation with a self-signed or verified -certificate, but please be aware that a valid certificate will be required in -Synapse v1.0. Instructions for having Synapse automatically provision and renew federation certificates through ACME can be found at [ACME.md](docs/ACME.md). - -If you would like to use your own certificates, you can do so by changing -`tls_certificate_path` and `tls_private_key_path` in `homeserver.yaml`; -alternatively, you can use a reverse proxy. See -[docs/reverse_proxy.rst](docs/reverse_proxy.rst) for information on configuring -a reverse proxy. +The default configuration exposes a single HTTP port: http://localhost:8008. It +is suitable for local testing, but for any practical use, you will either need +to enable a reverse proxy, or configure Synapse to expose an HTTPS port. + +For information on using a reverse proxy, see +[docs/reverse_proxy.rst](docs/reverse_proxy.rst). + +To configure Synapse to expose an HTTPS port, you will need to edit +`homeserver.yaml`. + +First, under the `listeners` section, uncomment the configuration for the +TLS-enabled listener. (Remove the hash sign (`#`) and space at the start of +each line). The relevant lines are like this: + +``` + - port: 8448 + type: http + tls: true + resources: + - names: [client, federation] +``` + +You will also need to uncomment the `tls_certificate_path` and +`tls_private_key_path` lines under the `TLS` section. You can either point +these settings at an existing certificate and key, or you can enable Synapse's +built-in ACME (Let's Encrypt) support. Instructions for having Synapse +automatically provision and renew federation certificates through ACME can be +found at [ACME.md](docs/ACME.md). ## Registering a user @@ -375,7 +391,7 @@ users. This can be done as follows: ``` $ source ~/synapse/env/bin/activate $ synctl start # if not already running -$ register_new_matrix_user -c homeserver.yaml https://localhost:8448 +$ register_new_matrix_user -c homeserver.yaml http://localhost:8008 New user localpart: erikj Password: Confirm password: diff --git a/README.rst b/README.rst index bc7cb5f784..9a7c04b55e 100644 --- a/README.rst +++ b/README.rst @@ -26,7 +26,6 @@ via IRC bridge at irc://irc.freenode.net/matrix. Synapse is currently in rapid development, but as of version 0.5 we believe it is sufficiently stable to be run as an internet-facing service for real usage! - About Matrix ============ @@ -88,18 +87,20 @@ Connecting to Synapse from a client =================================== The easiest way to try out your new Synapse installation is by connecting to it -from a web client. The easiest option is probably the one at -https://riot.im/app. You will need to specify a "Custom server" when you log on -or register: set this to ``https://domain.tld`` if you setup a reverse proxy -following the recommended setup, or ``https://localhost:8448`` - remember to specify the -port (``:8448``) if not ``:443`` unless you changed the configuration. (Leave the identity -server as the default - see `Identity servers`_.) - -If using port 8448 you will run into errors if you are using a self-signed -certificate. To overcome this, simply go to ``https://localhost:8448`` -directly with your browser and accept the presented certificate. You can then -go back in your web client and proceed further. Valid federation certificates -should not have this problem. +from a web client. + +Unless you are running a test instance of Synapse on your local machine, in +general, you will need to enable TLS support before you can successfully +connect from a client: see `<INSTALL.md#tls-certificates>`_. + +An easy way to get started is to login or register via Riot at +https://riot.im/app/#/login or https://riot.im/app/#/register respectively. +You will need to change the server you are logging into from ``matrix.org`` +and instead specify a Homeserver URL of ``https://<server_name>:8448`` +(or just ``https://<server_name>`` if you are using a reverse proxy). +(Leave the identity server as the default - see `Identity servers`_.) +If you prefer to use another client, refer to our +`client breakdown <https://matrix.org/docs/projects/clients-matrix>`_. If all goes well you should at least be able to log in, create a room, and start sending messages. @@ -174,9 +175,30 @@ Separately, Synapse may leak file handles if inbound HTTP requests get stuck during processing - e.g. blocked behind a lock or talking to a remote server etc. This is best diagnosed by matching up the 'Received request' and 'Processed request' log lines and looking for any 'Processed request' lines which take more than -a few seconds to execute. Please let us know at #matrix-dev:matrix.org if +a few seconds to execute. Please let us know at #synapse:matrix.org if you see this failure mode so we can help debug it, however. +Help!! Synapse eats all my RAM! +------------------------------- + +Synapse's architecture is quite RAM hungry currently - we deliberately +cache a lot of recent room data and metadata in RAM in order to speed up +common requests. We'll improve this in future, but for now the easiest +way to either reduce the RAM usage (at the risk of slowing things down) +is to set the almost-undocumented ``SYNAPSE_CACHE_FACTOR`` environment +variable. The default is 0.5, which can be decreased to reduce RAM usage +in memory constrained enviroments, or increased if performance starts to +degrade. + +Using `libjemalloc <http://jemalloc.net/>`_ can also yield a significant +improvement in overall amount, and especially in terms of giving back RAM +to the OS. To use it, the library must simply be put in the LD_PRELOAD +environment variable when launching Synapse. On Debian, this can be done +by installing the ``libjemalloc1`` package and adding this line to +``/etc/default/matrix-synapse``:: + + LD_PRELOAD=/usr/lib/x86_64-linux-gnu/libjemalloc.so.1 + Upgrading an existing Synapse ============================= @@ -196,12 +218,12 @@ Federation is the process by which users on different servers can participate in the same room. For this to work, those other servers must be able to contact yours to send messages. -The ``server_name`` in your -``homeserver.yaml`` file determines the way that other servers will reach -yours. By default, they will treat it as a hostname and try to connect to -port 8448. This is easy to set up and will work with the default configuration, -provided you set the ``server_name`` to match your machine's public DNS -hostname. +The ``server_name`` in your ``homeserver.yaml`` file determines the way that +other servers will reach yours. By default, they will treat it as a hostname +and try to connect to port 8448. This is easy to set up and will work with the +default configuration, provided you set the ``server_name`` to match your +machine's public DNS hostname, and give Synapse a TLS certificate which is +valid for your ``server_name``. For a more flexible configuration, you can set up a DNS SRV record. This allows you to run your server on a machine that might not have the same name as your @@ -243,11 +265,8 @@ largest boxes pause for thought.) Troubleshooting --------------- -You can use the federation tester to check if your homeserver is all set: -``https://matrix.org/federationtester/api/report?server_name=<your_server_name>`` -If any of the attributes under "checks" is false, federation won't work. -There is also a nicer interface available from a community member at -`<https://neo.lain.haus/fed-tester>`_. +You can use the `federation tester <https://matrix.org/federationtester>`_ to +check if your homeserver is all set. The typical failure mode with federation is that when you try to join a room, it is rejected with "401: Unauthorized". Generally this means that other @@ -263,7 +282,10 @@ So, things to check are: (it should be ``_matrix._tcp.<server_name>``), and that the port and hostname it specifies are reachable from outside your network. -.. TODO: add a note about forgetting ``nocanon`` on a reverse-proxy config +Another common problem is that people on other servers can't join rooms that +you invite them to. This can be caused by an incorrectly-configured reverse +proxy: see `<docs/reverse_proxy.rst>`_ for instructions on how to correctly +configure a reverse proxy. Running a Demo Federation of Synapses ------------------------------------- @@ -363,7 +385,7 @@ Synapse Development Before setting up a development environment for synapse, make sure you have the system dependencies (such as the python header files) installed - see -`Installing from source`_. +`Installing from source <INSTALL.md#installing-from-source>`_. To check out a synapse for development, clone the git repo into a working directory of your choice:: @@ -374,7 +396,7 @@ directory of your choice:: Synapse has a number of external dependencies, that are easiest to install using pip and a virtualenv:: - virtualenv -p python2.7 env + virtualenv -p python3 env source env/bin/activate python -m pip install -e .[all] @@ -416,25 +438,3 @@ sphinxcontrib-napoleon:: Building internal API documentation:: python setup.py build_sphinx - - -Help!! Synapse eats all my RAM! -=============================== - -Synapse's architecture is quite RAM hungry currently - we deliberately -cache a lot of recent room data and metadata in RAM in order to speed up -common requests. We'll improve this in future, but for now the easiest -way to either reduce the RAM usage (at the risk of slowing things down) -is to set the almost-undocumented ``SYNAPSE_CACHE_FACTOR`` environment -variable. The default is 0.5, which can be decreased to reduce RAM usage -in memory constrained enviroments, or increased if performance starts to -degrade. - -Using `libjemalloc <http://jemalloc.net/>`_ can also yield a significant -improvement in overall amount, and especially in terms of giving back RAM -to the OS. To use it, the library must simply be put in the LD_PRELOAD -environment variable when launching Synapse. On Debian, this can be done -by installing the ``libjemalloc1`` package and adding this line to -``/etc/default/matrix-synapse``:: - - LD_PRELOAD=/usr/lib/x86_64-linux-gnu/libjemalloc.so.1 diff --git a/changelog.d/4614.feature b/changelog.d/4614.feature new file mode 100644 index 0000000000..18e16dbc7b --- /dev/null +++ b/changelog.d/4614.feature @@ -0,0 +1 @@ +The default configuration no longer requires TLS certificates. diff --git a/changelog.d/4618.bugfix b/changelog.d/4618.bugfix new file mode 100644 index 0000000000..22115a020e --- /dev/null +++ b/changelog.d/4618.bugfix @@ -0,0 +1 @@ +Fix failure to start when not TLS certificate was given even if TLS was disabled. diff --git a/changelog.d/4621.misc b/changelog.d/4621.misc new file mode 100644 index 0000000000..60e45cb70c --- /dev/null +++ b/changelog.d/4621.misc @@ -0,0 +1 @@ +README updates diff --git a/changelog.d/4625.bugfix b/changelog.d/4625.bugfix new file mode 100644 index 0000000000..3dc0ecf24c --- /dev/null +++ b/changelog.d/4625.bugfix @@ -0,0 +1 @@ +fix self-signed cert notice from generate-config. diff --git a/changelog.d/4626.misc b/changelog.d/4626.misc new file mode 100644 index 0000000000..f1a57dcf9a --- /dev/null +++ b/changelog.d/4626.misc @@ -0,0 +1 @@ +Improve 'user_ips' table deduplication background update diff --git a/synapse/app/_base.py b/synapse/app/_base.py index 50fd17c0be..5b0ca312e2 100644 --- a/synapse/app/_base.py +++ b/synapse/app/_base.py @@ -213,12 +213,13 @@ def refresh_certificate(hs): Refresh the TLS certificates that Synapse is using by re-reading them from disk and updating the TLS context factories to use them. """ - hs.config.read_certificate_from_disk() if not hs.config.has_tls_listener(): - # nothing else to do here + # attempt to reload the certs for the good of the tls_fingerprints + hs.config.read_certificate_from_disk(require_cert_and_key=False) return + hs.config.read_certificate_from_disk(require_cert_and_key=True) hs.tls_server_context_factory = context_factory.ServerContextFactory(hs.config) if hs._listening_services: diff --git a/synapse/config/_base.py b/synapse/config/_base.py index 5858fb92b4..5aec43b702 100644 --- a/synapse/config/_base.py +++ b/synapse/config/_base.py @@ -257,7 +257,7 @@ class Config(object): "--keys-directory", metavar="DIRECTORY", help="Used with 'generate-*' options to specify where files such as" - " certs and signing keys should be stored in, unless explicitly" + " signing keys should be stored, unless explicitly" " specified in the config.", ) config_parser.add_argument( @@ -313,16 +313,11 @@ class Config(object): print( ( "A config file has been generated in %r for server name" - " %r with corresponding SSL keys and self-signed" - " certificates. Please review this file and customise it" + " %r. Please review this file and customise it" " to your needs." ) % (config_path, server_name) ) - print( - "If this server name is incorrect, you will need to" - " regenerate the SSL certificates" - ) return else: print( diff --git a/synapse/config/server.py b/synapse/config/server.py index 767897c419..c5c3aac8ed 100644 --- a/synapse/config/server.py +++ b/synapse/config/server.py @@ -387,47 +387,47 @@ class ServerConfig(Config): # webclient: A web client. Requires web_client_location to be set. # listeners: - # Main HTTPS listener. - # For when matrix traffic is sent directly to synapse. - - port: %(bind_port)s + # TLS-enabled listener: for when matrix traffic is sent directly to synapse. + # + # Disabled by default. To enable it, uncomment the following. (Note that you + # will also need to give Synapse a TLS key and certificate: see the TLS section + # below.) + # + # - port: %(bind_port)s + # type: http + # tls: true + # resources: + # - names: [client, federation] + + # Unsecure HTTP listener: for when matrix traffic passes through a reverse proxy + # that unwraps TLS. + # + # If you plan to use a reverse proxy, please see + # https://github.com/matrix-org/synapse/blob/master/docs/reverse_proxy.rst. + # + - port: %(unsecure_port)s + tls: false + bind_addresses: ['::1', '127.0.0.1'] type: http - tls: true + x_forwarded: true - # List of HTTP resources to serve on this listener. resources: - - names: [client] - compress: true - - names: [federation] + - names: [client, federation] compress: false - # example addional_resources: + # example additonal_resources: # # additional_resources: # "/_matrix/my/custom/endpoint": # module: my_module.CustomRequestHandler # config: {} - # Unsecure HTTP listener - # For when matrix traffic passes through a reverse-proxy that unwraps TLS. - - port: %(unsecure_port)s - tls: false - bind_addresses: ['::1', '127.0.0.1'] - type: http - x_forwarded: true - - resources: - - names: [client] - compress: true - - names: [federation] - compress: false - # Turn on the twisted ssh manhole service on localhost on the given # port. # - port: 9000 # bind_addresses: ['::1', '127.0.0.1'] # type: manhole - # Homeserver blocking # # How to reach the server admin, used in ResourceLimitError diff --git a/synapse/config/tls.py b/synapse/config/tls.py index 86e6eb80db..5fb3486db1 100644 --- a/synapse/config/tls.py +++ b/synapse/config/tls.py @@ -23,7 +23,7 @@ from unpaddedbase64 import encode_base64 from OpenSSL import crypto -from synapse.config._base import Config +from synapse.config._base import Config, ConfigError logger = logging.getLogger(__name__) @@ -45,6 +45,19 @@ class TlsConfig(Config): self.tls_certificate_file = self.abspath(config.get("tls_certificate_path")) self.tls_private_key_file = self.abspath(config.get("tls_private_key_path")) + + if self.has_tls_listener(): + if not self.tls_certificate_file: + raise ConfigError( + "tls_certificate_path must be specified if TLS-enabled listeners are " + "configured." + ) + if not self.tls_private_key_file: + raise ConfigError( + "tls_certificate_path must be specified if TLS-enabled listeners are " + "configured." + ) + self._original_tls_fingerprints = config.get("tls_fingerprints", []) if self._original_tls_fingerprints is None: @@ -105,26 +118,40 @@ class TlsConfig(Config): days_remaining = (expires_on - now).days return days_remaining - def read_certificate_from_disk(self): - """ - Read the certificates from disk. + def read_certificate_from_disk(self, require_cert_and_key): """ - self.tls_certificate = self.read_tls_certificate() + Read the certificates and private key from disk. - if self.has_tls_listener(): + Args: + require_cert_and_key (bool): set to True to throw an error if the certificate + and key file are not given + """ + if require_cert_and_key: self.tls_private_key = self.read_tls_private_key() + self.tls_certificate = self.read_tls_certificate() + elif self.tls_certificate_file: + # we only need the certificate for the tls_fingerprints. Reload it if we + # can, but it's not a fatal error if we can't. + try: + self.tls_certificate = self.read_tls_certificate() + except Exception as e: + logger.info( + "Unable to read TLS certificate (%s). Ignoring as no " + "tls listeners enabled.", e, + ) self.tls_fingerprints = list(self._original_tls_fingerprints) - # Check that our own certificate is included in the list of fingerprints - # and include it if it is not. - x509_certificate_bytes = crypto.dump_certificate( - crypto.FILETYPE_ASN1, self.tls_certificate - ) - sha256_fingerprint = encode_base64(sha256(x509_certificate_bytes).digest()) - sha256_fingerprints = set(f["sha256"] for f in self.tls_fingerprints) - if sha256_fingerprint not in sha256_fingerprints: - self.tls_fingerprints.append({u"sha256": sha256_fingerprint}) + if self.tls_certificate: + # Check that our own certificate is included in the list of fingerprints + # and include it if it is not. + x509_certificate_bytes = crypto.dump_certificate( + crypto.FILETYPE_ASN1, self.tls_certificate + ) + sha256_fingerprint = encode_base64(sha256(x509_certificate_bytes).digest()) + sha256_fingerprints = set(f["sha256"] for f in self.tls_fingerprints) + if sha256_fingerprint not in sha256_fingerprints: + self.tls_fingerprints.append({u"sha256": sha256_fingerprint}) def default_config(self, config_dir_path, server_name, **kwargs): base_key_name = os.path.join(config_dir_path, server_name) @@ -149,10 +176,10 @@ class TlsConfig(Config): # See 'ACME support' below to enable auto-provisioning this certificate via # Let's Encrypt. # - tls_certificate_path: "%(tls_certificate_path)s" + # tls_certificate_path: "%(tls_certificate_path)s" # PEM-encoded private key for TLS - tls_private_key_path: "%(tls_private_key_path)s" + # tls_private_key_path: "%(tls_private_key_path)s" # ACME support: This will configure Synapse to request a valid TLS certificate # for your configured `server_name` via Let's Encrypt. @@ -177,7 +204,7 @@ class TlsConfig(Config): # acme: # ACME support is disabled by default. Uncomment the following line - # to enable it. + # (and tls_certificate_path and tls_private_key_path above) to enable it. # # enabled: true diff --git a/synapse/storage/client_ips.py b/synapse/storage/client_ips.py index cc23d7cdbe..9c21362226 100644 --- a/synapse/storage/client_ips.py +++ b/synapse/storage/client_ips.py @@ -191,12 +191,16 @@ class ClientIpStore(background_updates.BackgroundUpdateStore): clause = "? <= last_seen AND last_seen < ?" args = (begin_last_seen, end_last_seen) + # (Note: The DISTINCT in the inner query is important to ensure that + # the COUNT(*) is accurate, otherwise double counting may happen due + # to the join effectively being a cross product) txn.execute( """ SELECT user_id, access_token, ip, - MAX(device_id), MAX(user_agent), MAX(last_seen) + MAX(device_id), MAX(user_agent), MAX(last_seen), + COUNT(*) FROM ( - SELECT user_id, access_token, ip + SELECT DISTINCT user_id, access_token, ip FROM user_ips WHERE {} ) c @@ -210,7 +214,60 @@ class ClientIpStore(background_updates.BackgroundUpdateStore): # We've got some duplicates for i in res: - user_id, access_token, ip, device_id, user_agent, last_seen = i + user_id, access_token, ip, device_id, user_agent, last_seen, count = i + + # We want to delete the duplicates so we end up with only a + # single row. + # + # The naive way of doing this would be just to delete all rows + # and reinsert a constructed row. However, if there are a lot of + # duplicate rows this can cause the table to grow a lot, which + # can be problematic in two ways: + # 1. If user_ips is already large then this can cause the + # table to rapidly grow, potentially filling the disk. + # 2. Reinserting a lot of rows can confuse the table + # statistics for postgres, causing it to not use the + # correct indices for the query above, resulting in a full + # table scan. This is incredibly slow for large tables and + # can kill database performance. (This seems to mainly + # happen for the last query where the clause is simply `? < + # last_seen`) + # + # So instead we want to delete all but *one* of the duplicate + # rows. That is hard to do reliably, so we cheat and do a two + # step process: + # 1. Delete all rows with a last_seen strictly less than the + # max last_seen. This hopefully results in deleting all but + # one row the majority of the time, but there may be + # duplicate last_seen + # 2. If multiple rows remain, we fall back to the naive method + # and simply delete all rows and reinsert. + # + # Note that this relies on no new duplicate rows being inserted, + # but if that is happening then this entire process is futile + # anyway. + + # Do step 1: + + txn.execute( + """ + DELETE FROM user_ips + WHERE user_id = ? AND access_token = ? AND ip = ? AND last_seen < ? + """, + (user_id, access_token, ip, last_seen) + ) + if txn.rowcount == count - 1: + # We deleted all but one of the duplicate rows, i.e. there + # is exactly one remaining and so there is nothing left to + # do. + continue + elif txn.rowcount >= count: + raise Exception( + "We deleted more duplicate rows from 'user_ips' than expected", + ) + + # The previous step didn't delete enough rows, so we fallback to + # step 2: # Drop all the duplicates txn.execute( diff --git a/tests/config/test_tls.py b/tests/config/test_tls.py index d8fd18a9cb..c260d3359f 100644 --- a/tests/config/test_tls.py +++ b/tests/config/test_tls.py @@ -65,7 +65,7 @@ s4niecZKPBizL6aucT59CsunNmmb5Glq8rlAcU+1ZTZZzGYqVYhF6axB9Qg= t = TestConfig() t.read_config(config) - t.read_certificate_from_disk() + t.read_certificate_from_disk(require_cert_and_key=False) warnings = self.flushWarnings() self.assertEqual(len(warnings), 1) |