synapse/tests/replication/_base.py

# Copyright 2019 New Vector Ltd
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import logging
from collections import defaultdict
from typing import Any, Dict, List, Optional, Set, Tuple

from twisted.internet.address import IPv4Address
from twisted.internet.protocol import Protocol, connectionDone
from twisted.python.failure import Failure
from twisted.test.proto_helpers import MemoryReactor
from twisted.web.resource import Resource

from synapse.app.generic_worker import GenericWorkerServer
from synapse.http.site import SynapseRequest, SynapseSite
from synapse.replication.http import ReplicationRestResource
from synapse.replication.tcp.client import ReplicationDataHandler
from synapse.replication.tcp.handler import ReplicationCommandHandler
from synapse.replication.tcp.protocol import (
    ClientReplicationStreamProtocol,
    ServerReplicationStreamProtocol,
)
from synapse.replication.tcp.resource import ReplicationStreamProtocolFactory
from synapse.server import HomeServer
from synapse.util import Clock

from tests import unittest
from tests.server import FakeTransport
from tests.utils import USE_POSTGRES_FOR_TESTS

try:
    import hiredis
except ImportError:
    hiredis = None  # type: ignore

logger = logging.getLogger(__name__)


class BaseStreamTestCase(unittest.HomeserverTestCase):
    """Base class for tests of the replication streams"""

    # hiredis is an optional dependency so we don't want to require it for running
    # the tests.
    if not hiredis:
        skip = "Requires hiredis"

    if not USE_POSTGRES_FOR_TESTS:
        # Redis replication only takes place on Postgres
        skip = "Requires Postgres"

    def prepare(self, reactor: MemoryReactor, clock: Clock, hs: HomeServer) -> None:
        # build a replication server
        server_factory = ReplicationStreamProtocolFactory(hs)
        self.streamer = hs.get_replication_streamer()
        self.server: ServerReplicationStreamProtocol = server_factory.buildProtocol(
            IPv4Address("TCP", "127.0.0.1", 0)
        )

        # Make a new HomeServer object for the worker
        self.reactor.lookups["testserv"] = "1.2.3.4"
        self.worker_hs = self.setup_test_homeserver(
            federation_http_client=None,
            homeserver_to_use=GenericWorkerServer,
            config=self._get_worker_hs_config(),
            reactor=self.reactor,
        )

        # Since we use sqlite in memory databases we need to make sure the
        # databases objects are the same.
        self.worker_hs.get_datastores().main.db_pool = hs.get_datastores().main.db_pool

        # Normally we'd pass in the handler to `setup_test_homeserver`, which would
        # eventually hit "Install @cache_in_self attributes" in tests/utils.py.
        # Unfortunately our handler wants a reference to the homeserver. That leaves
        # us with a chicken-and-egg problem.
        # We can workaround this: create the homeserver first, create the handler
        # and bodge it in after the fact. The bodging requires us to know the
        # dirty details of how `cache_in_self` works. We politely ask mypy to
        # ignore our dirty dealings.
        self.test_handler = self._build_replication_data_handler()
        self.worker_hs._replication_data_handler = self.test_handler  # type: ignore[attr-defined]

        repl_handler = ReplicationCommandHandler(self.worker_hs)
        self.client = ClientReplicationStreamProtocol(
            self.worker_hs,
            "client",
            "test",
            clock,
            repl_handler,
        )

        self._client_transport: Optional[FakeTransport] = None
        self._server_transport: Optional[FakeTransport] = None

    def create_resource_dict(self) -> Dict[str, Resource]:
        d = super().create_resource_dict()
        d["/_synapse/replication"] = ReplicationRestResource(self.hs)
        return d

    def _get_worker_hs_config(self) -> dict:
        config = self.default_config()
        config["worker_app"] = "synapse.app.generic_worker"
        config["instance_map"] = {"main": {"host": "testserv", "port": 8765}}
        return config

    def _build_replication_data_handler(self) -> "TestReplicationDataHandler":
        return TestReplicationDataHandler(self.worker_hs)

    def reconnect(self) -> None:
        if self._client_transport:
            self.client.close()

        if self._server_transport:
            self.server.close()

        self._client_transport = FakeTransport(self.server, self.reactor)
        self.client.makeConnection(self._client_transport)

        self._server_transport = FakeTransport(self.client, self.reactor)
        self.server.makeConnection(self._server_transport)

    def disconnect(self) -> None:
        if self._client_transport:
            self._client_transport = None
            self.client.close()

        if self._server_transport:
            self._server_transport = None
            self.server.close()

    def replicate(self) -> None:
        """Tell the master side of replication that something has happened, and then
        wait for the replication to occur.
        """
        self.streamer.on_notifier_poke()
        self.pump(0.1)

    def handle_http_replication_attempt(self) -> SynapseRequest:
        """Asserts that a connection attempt was made to the master HS on the
        HTTP replication port, then proxies it to the master HS object to be
        handled.

        Returns:
            The request object received by master HS.
        """

        # We should have an outbound connection attempt.
        clients = self.reactor.tcpClients
        self.assertEqual(len(clients), 1)
        (host, port, client_factory, _timeout, _bindAddress) = clients.pop(0)
        self.assertEqual(host, "1.2.3.4")
        self.assertEqual(port, 8765)

        # Set up client side protocol
        client_address = IPv4Address("TCP", "127.0.0.1", 1234)
        client_protocol = client_factory.buildProtocol(("127.0.0.1", 1234))

        # Set up the server side protocol
        server_address = IPv4Address("TCP", host, port)
        channel = self.site.buildProtocol((host, port))

        # hook into the channel's request factory so that we can keep a record
        # of the requests
        requests: List[SynapseRequest] = []
        real_request_factory = channel.requestFactory

        def request_factory(*args: Any, **kwargs: Any) -> SynapseRequest:
            request = real_request_factory(*args, **kwargs)
            requests.append(request)
            return request

        channel.requestFactory = request_factory

        # Connect client to server and vice versa.
        client_to_server_transport = FakeTransport(
            channel, self.reactor, client_protocol, server_address, client_address
        )
        client_protocol.makeConnection(client_to_server_transport)

        server_to_client_transport = FakeTransport(
            client_protocol, self.reactor, channel, client_address, server_address
        )
        channel.makeConnection(server_to_client_transport)

        # The request will now be processed by `self.site` and the response
        # streamed back.
        self.reactor.advance(0)

        # We tear down the connection so it doesn't get reused without our
        # knowledge.
        server_to_client_transport.loseConnection()
        client_to_server_transport.loseConnection()

        # there should have been exactly one request
        self.assertEqual(len(requests), 1)

        return requests[0]

    def assert_request_is_get_repl_stream_updates(
        self, request: SynapseRequest, stream_name: str
    ) -> None:
        """Asserts that the given request is a HTTP replication request for
        fetching updates for given stream.
        """

        path: bytes = request.path  # type: ignore
        self.assertRegex(
            path,
            rb"^/_synapse/replication/get_repl_stream_updates/%s/[^/]+$"
            % (stream_name.encode("ascii"),),
        )

        self.assertEqual(request.method, b"GET")


class BaseMultiWorkerStreamTestCase(unittest.HomeserverTestCase):
    """Base class for tests running multiple workers.

    Enables Redis, providing a fake Redis server.

    Automatically handle HTTP replication requests from workers to master,
    unlike `BaseStreamTestCase`.
    """

    if not hiredis:
        skip = "Requires hiredis"

    if not USE_POSTGRES_FOR_TESTS:
        # Redis replication only takes place on Postgres
        skip = "Requires Postgres"

    def default_config(self) -> Dict[str, Any]:
        """
        Overrides the default config to enable Redis.
        Even if the test only uses make_worker_hs, the main process needs Redis
        enabled otherwise it won't create a Fake Redis server to listen on the
        Redis port and accept fake TCP connections.
        """
        base = super().default_config()
        base["redis"] = {"enabled": True}
        base["instance_map"] = {"main": {"host": "testserv", "port": 8765}}
        return base

    def setUp(self) -> None:
        super().setUp()

        # build a replication server
        self.streamer = self.hs.get_replication_streamer()

        # Fake in memory Redis server that servers can connect to.
        self._redis_server = FakeRedisPubSubServer()

        # We may have an attempt to connect to redis for the external cache already.
        self.connect_any_redis_attempts()

        store = self.hs.get_datastores().main
        self.database_pool = store.db_pool

        self.reactor.lookups["testserv"] = "1.2.3.4"
        self.reactor.lookups["localhost"] = "127.0.0.1"

        # A map from a HS instance to the associated HTTP Site to use for
        # handling inbound HTTP requests to that instance.
        self._hs_to_site = {self.hs: self.site}

        # Handle attempts to connect to fake redis server.
        self.reactor.add_tcp_client_callback(
            "localhost",
            6379,
            self.connect_any_redis_attempts,
        )

        self.hs.get_replication_command_handler().start_replication(self.hs)

        # When we see a connection attempt to the master replication listener we
        # automatically set up the connection. This is so that tests don't
        # manually have to go and explicitly set it up each time (plus sometimes
        # it is impossible to write the handling explicitly in the tests).
        #
        # Register the master replication listener:
        self.reactor.add_tcp_client_callback(
            "1.2.3.4",
            8765,
            lambda: self._handle_http_replication_attempt(self.hs, 8765),
        )

    def create_test_resource(self) -> ReplicationRestResource:
        """Overrides `HomeserverTestCase.create_test_resource`."""
        # We override this so that it automatically registers all the HTTP
        # replication servlets, without having to explicitly do that in all
        # subclassses.

        resource = ReplicationRestResource(self.hs)

        for servlet in self.servlets:
            servlet(self.hs, resource)

        return resource

    def make_worker_hs(
        self, worker_app: str, extra_config: Optional[dict] = None, **kwargs: Any
    ) -> HomeServer:
        """Make a new worker HS instance, correctly connecting replication
        stream to the master HS.

        Args:
            worker_app: Type of worker, e.g. `synapse.app.generic_worker`.
            extra_config: Any extra config to use for this instances.
            **kwargs: Options that get passed to `self.setup_test_homeserver`,
                useful to e.g. pass some mocks for things like `federation_http_client`

        Returns:
            The new worker HomeServer instance.
        """

        config = self._get_worker_hs_config()
        config["worker_app"] = worker_app
        config.update(extra_config or {})

        worker_hs = self.setup_test_homeserver(
            homeserver_to_use=GenericWorkerServer,
            config=config,
            reactor=self.reactor,
            **kwargs,
        )

        # If the instance is in the `instance_map` config then workers may try
        # and send HTTP requests to it, so we register it with
        # `_handle_http_replication_attempt` like we do with the master HS.
        instance_name = worker_hs.get_instance_name()
        instance_loc = worker_hs.config.worker.instance_map.get(instance_name)
        if instance_loc:
            # Ensure the host is one that has a fake DNS entry.
            if instance_loc.host not in self.reactor.lookups:
                raise Exception(
                    "Host does not have an IP for instance_map[%r].host = %r"
                    % (
                        instance_name,
                        instance_loc.host,
                    )
                )

            # Copy the port into a new, non-Optional variable so mypy knows we're
            # not going to reset `instance_loc` to `None` under its feet. See
            # https://mypy.readthedocs.io/en/latest/common_issues.html#narrowing-and-inner-functions
            port = instance_loc.port

            self.reactor.add_tcp_client_callback(
                self.reactor.lookups[instance_loc.host],
                instance_loc.port,
                lambda: self._handle_http_replication_attempt(worker_hs, port),
            )

        store = worker_hs.get_datastores().main
        store.db_pool._db_pool = self.database_pool._db_pool

        # Set up a resource for the worker
        resource = ReplicationRestResource(worker_hs)

        for servlet in self.servlets:
            servlet(worker_hs, resource)

        self._hs_to_site[worker_hs] = SynapseSite(
            logger_name="synapse.access.http.fake",
            site_tag="{}-{}".format(
                worker_hs.config.server.server_name, worker_hs.get_instance_name()
            ),
            config=worker_hs.config.server.listeners[0],
            resource=resource,
            server_version_string="1",
            max_request_body_size=8192,
            reactor=self.reactor,
        )

        worker_hs.get_replication_command_handler().start_replication(worker_hs)

        return worker_hs

    def _get_worker_hs_config(self) -> dict:
        config = self.default_config()
        return config

    def replicate(self) -> None:
        """Tell the master side of replication that something has happened, and then
        wait for the replication to occur.
        """
        self.streamer.on_notifier_poke()
        self.pump()

    def _handle_http_replication_attempt(self, hs: HomeServer, repl_port: int) -> None:
        """Handles a connection attempt to the given HS replication HTTP
        listener on the given port.
        """

        # We should have at least one outbound connection attempt, where the
        # last is one to the HTTP repication IP/port.
        clients = self.reactor.tcpClients
        self.assertGreaterEqual(len(clients), 1)
        (host, port, client_factory, _timeout, _bindAddress) = clients.pop()
        self.assertEqual(host, "1.2.3.4")
        self.assertEqual(port, repl_port)

        # Set up client side protocol
        client_address = IPv4Address("TCP", "127.0.0.1", 1234)
        client_protocol = client_factory.buildProtocol(("127.0.0.1", 1234))

        # Set up the server side protocol
        server_address = IPv4Address("TCP", host, port)
        channel = self._hs_to_site[hs].buildProtocol((host, port))

        # Connect client to server and vice versa.
        client_to_server_transport = FakeTransport(
            channel, self.reactor, client_protocol, server_address, client_address
        )
        client_protocol.makeConnection(client_to_server_transport)

        server_to_client_transport = FakeTransport(
            client_protocol, self.reactor, channel, client_address, server_address
        )
        channel.makeConnection(server_to_client_transport)

        # Note: at this point we've wired everything up, but we need to return
        # before the data starts flowing over the connections as this is called
        # inside `connecTCP` before the connection has been passed back to the
        # code that requested the TCP connection.

    def connect_any_redis_attempts(self) -> None:
        """If redis is enabled we need to deal with workers connecting to a
        redis server. We don't want to use a real Redis server so we use a
        fake one.
        """
        clients = self.reactor.tcpClients
        while clients:
            (host, port, client_factory, _timeout, _bindAddress) = clients.pop(0)
            self.assertEqual(host, "localhost")
            self.assertEqual(port, 6379)

            client_address = IPv4Address("TCP", "127.0.0.1", 6379)
            client_protocol = client_factory.buildProtocol(client_address)

            server_address = IPv4Address("TCP", host, port)
            server_protocol = self._redis_server.buildProtocol(server_address)

            client_to_server_transport = FakeTransport(
                server_protocol, self.reactor, client_protocol
            )
            client_protocol.makeConnection(client_to_server_transport)

            server_to_client_transport = FakeTransport(
                client_protocol, self.reactor, server_protocol
            )
            server_protocol.makeConnection(server_to_client_transport)


class TestReplicationDataHandler(ReplicationDataHandler):
    """Drop-in for ReplicationDataHandler which just collects RDATA rows"""

    def __init__(self, hs: HomeServer):
        super().__init__(hs)

        # list of received (stream_name, token, row) tuples
        self.received_rdata_rows: List[Tuple[str, int, Any]] = []

    async def on_rdata(
        self, stream_name: str, instance_name: str, token: int, rows: list
    ) -> None:
        await super().on_rdata(stream_name, instance_name, token, rows)
        for r in rows:
            self.received_rdata_rows.append((stream_name, token, r))


class FakeRedisPubSubServer:
    """A fake Redis server for pub/sub."""

    def __init__(self) -> None:
        self._subscribers_by_channel: Dict[
            bytes, Set["FakeRedisPubSubProtocol"]
        ] = defaultdict(set)

    def add_subscriber(self, conn: "FakeRedisPubSubProtocol", channel: bytes) -> None:
        """A connection has called SUBSCRIBE"""
        self._subscribers_by_channel[channel].add(conn)

    def remove_subscriber(self, conn: "FakeRedisPubSubProtocol") -> None:
        """A connection has lost connection"""
        for subscribers in self._subscribers_by_channel.values():
            subscribers.discard(conn)

    def publish(
        self, conn: "FakeRedisPubSubProtocol", channel: bytes, msg: object
    ) -> int:
        """A connection want to publish a message to subscribers."""
        for sub in self._subscribers_by_channel[channel]:
            sub.send(["message", channel, msg])

        return len(self._subscribers_by_channel)

    def buildProtocol(self, addr: IPv4Address) -> "FakeRedisPubSubProtocol":
        return FakeRedisPubSubProtocol(self)


class FakeRedisPubSubProtocol(Protocol):
    """A connection from a client talking to the fake Redis server."""

    transport: Optional[FakeTransport] = None

    def __init__(self, server: FakeRedisPubSubServer):
        self._server = server
        self._reader = hiredis.Reader()

    def dataReceived(self, data: bytes) -> None:
        self._reader.feed(data)

        # We might get multiple messages in one packet.
        while True:
            msg = self._reader.gets()

            if msg is False:
                # No more messages.
                return

            if not isinstance(msg, list):
                # Inbound commands should always be a list
                raise Exception("Expected redis list")

            self.handle_command(msg[0], *msg[1:])

    def handle_command(self, command: bytes, *args: bytes) -> None:
        """Received a Redis command from the client."""

        # We currently only support pub/sub.
        if command == b"PUBLISH":
            channel, message = args
            num_subscribers = self._server.publish(self, channel, message)
            self.send(num_subscribers)
        elif command == b"SUBSCRIBE":
            for idx, channel in enumerate(args):
                num_channels = idx + 1
                self._server.add_subscriber(self, channel)
                self.send(["subscribe", channel, num_channels])

        # Since we use SET/GET to cache things we can safely no-op them.
        elif command == b"SET":
            self.send("OK")
        elif command == b"GET":
            self.send(None)

        # Connection keep-alives.
        elif command == b"PING":
            self.send("PONG")

        else:
            raise Exception(f"Unknown command: {command!r}")

    def send(self, msg: object) -> None:
        """Send a message back to the client."""
        assert self.transport is not None

        raw = self.encode(msg).encode("utf-8")

        self.transport.write(raw)
        self.transport.flush()

    def encode(self, obj: object) -> str:
        """Encode an object to its Redis format.

        Supports: strings/bytes, integers and list/tuples.
        """

        if isinstance(obj, bytes):
            # We assume bytes are just unicode strings.
            obj = obj.decode("utf-8")

        if obj is None:
            return "$-1\r\n"
        if isinstance(obj, str):
            return f"${len(obj)}\r\n{obj}\r\n"
        if isinstance(obj, int):
            return f":{obj}\r\n"
        if isinstance(obj, (list, tuple)):
            items = "".join(self.encode(a) for a in obj)
            return f"*{len(obj)}\r\n{items}"

        raise Exception("Unrecognized type for encoding redis: %r: %r", type(obj), obj)

    def connectionLost(self, reason: Failure = connectionDone) -> None:
        self._server.remove_subscriber(self)