mirror of
https://mau.dev/maunium/synapse.git
synced 2024-10-01 01:36:05 -04:00
Add ability to run multiple pusher instances (#7855)
This reuses the same scheme as federation sender sharding
This commit is contained in:
parent
a827838706
commit
649a7ead5c
1
changelog.d/7855.feature
Normal file
1
changelog.d/7855.feature
Normal file
@ -0,0 +1 @@
|
|||||||
|
Add experimental support for running multiple pusher workers.
|
@ -19,9 +19,11 @@ import argparse
|
|||||||
import errno
|
import errno
|
||||||
import os
|
import os
|
||||||
from collections import OrderedDict
|
from collections import OrderedDict
|
||||||
|
from hashlib import sha256
|
||||||
from textwrap import dedent
|
from textwrap import dedent
|
||||||
from typing import Any, MutableMapping, Optional
|
from typing import Any, List, MutableMapping, Optional
|
||||||
|
|
||||||
|
import attr
|
||||||
import yaml
|
import yaml
|
||||||
|
|
||||||
|
|
||||||
@ -717,4 +719,36 @@ def find_config_files(search_paths):
|
|||||||
return config_files
|
return config_files
|
||||||
|
|
||||||
|
|
||||||
__all__ = ["Config", "RootConfig"]
|
@attr.s
|
||||||
|
class ShardedWorkerHandlingConfig:
|
||||||
|
"""Algorithm for choosing which instance is responsible for handling some
|
||||||
|
sharded work.
|
||||||
|
|
||||||
|
For example, the federation senders use this to determine which instances
|
||||||
|
handles sending stuff to a given destination (which is used as the `key`
|
||||||
|
below).
|
||||||
|
"""
|
||||||
|
|
||||||
|
instances = attr.ib(type=List[str])
|
||||||
|
|
||||||
|
def should_handle(self, instance_name: str, key: str) -> bool:
|
||||||
|
"""Whether this instance is responsible for handling the given key.
|
||||||
|
"""
|
||||||
|
|
||||||
|
# If multiple instances are not defined we always return true.
|
||||||
|
if not self.instances or len(self.instances) == 1:
|
||||||
|
return True
|
||||||
|
|
||||||
|
# We shard by taking the hash, modulo it by the number of instances and
|
||||||
|
# then checking whether this instance matches the instance at that
|
||||||
|
# index.
|
||||||
|
#
|
||||||
|
# (Technically this introduces some bias and is not entirely uniform,
|
||||||
|
# but since the hash is so large the bias is ridiculously small).
|
||||||
|
dest_hash = sha256(key.encode("utf8")).digest()
|
||||||
|
dest_int = int.from_bytes(dest_hash, byteorder="little")
|
||||||
|
remainder = dest_int % (len(self.instances))
|
||||||
|
return self.instances[remainder] == instance_name
|
||||||
|
|
||||||
|
|
||||||
|
__all__ = ["Config", "RootConfig", "ShardedWorkerHandlingConfig"]
|
||||||
|
@ -137,3 +137,8 @@ class Config:
|
|||||||
|
|
||||||
def read_config_files(config_files: List[str]): ...
|
def read_config_files(config_files: List[str]): ...
|
||||||
def find_config_files(search_paths: List[str]): ...
|
def find_config_files(search_paths: List[str]): ...
|
||||||
|
|
||||||
|
class ShardedWorkerHandlingConfig:
|
||||||
|
instances: List[str]
|
||||||
|
def __init__(self, instances: List[str]) -> None: ...
|
||||||
|
def should_handle(self, instance_name: str, key: str) -> bool: ...
|
||||||
|
@ -13,42 +13,11 @@
|
|||||||
# See the License for the specific language governing permissions and
|
# See the License for the specific language governing permissions and
|
||||||
# limitations under the License.
|
# limitations under the License.
|
||||||
|
|
||||||
from hashlib import sha256
|
from typing import Optional
|
||||||
from typing import List, Optional
|
|
||||||
|
|
||||||
import attr
|
|
||||||
from netaddr import IPSet
|
from netaddr import IPSet
|
||||||
|
|
||||||
from ._base import Config, ConfigError
|
from ._base import Config, ConfigError, ShardedWorkerHandlingConfig
|
||||||
|
|
||||||
|
|
||||||
@attr.s
|
|
||||||
class ShardedFederationSendingConfig:
|
|
||||||
"""Algorithm for choosing which federation sender instance is responsible
|
|
||||||
for which destionation host.
|
|
||||||
"""
|
|
||||||
|
|
||||||
instances = attr.ib(type=List[str])
|
|
||||||
|
|
||||||
def should_send_to(self, instance_name: str, destination: str) -> bool:
|
|
||||||
"""Whether this instance is responsible for sending transcations for
|
|
||||||
the given host.
|
|
||||||
"""
|
|
||||||
|
|
||||||
# If multiple federation senders are not defined we always return true.
|
|
||||||
if not self.instances or len(self.instances) == 1:
|
|
||||||
return True
|
|
||||||
|
|
||||||
# We shard by taking the hash, modulo it by the number of federation
|
|
||||||
# senders and then checking whether this instance matches the instance
|
|
||||||
# at that index.
|
|
||||||
#
|
|
||||||
# (Technically this introduces some bias and is not entirely uniform, but
|
|
||||||
# since the hash is so large the bias is ridiculously small).
|
|
||||||
dest_hash = sha256(destination.encode("utf8")).digest()
|
|
||||||
dest_int = int.from_bytes(dest_hash, byteorder="little")
|
|
||||||
remainder = dest_int % (len(self.instances))
|
|
||||||
return self.instances[remainder] == instance_name
|
|
||||||
|
|
||||||
|
|
||||||
class FederationConfig(Config):
|
class FederationConfig(Config):
|
||||||
@ -61,7 +30,7 @@ class FederationConfig(Config):
|
|||||||
self.send_federation = config.get("send_federation", True)
|
self.send_federation = config.get("send_federation", True)
|
||||||
|
|
||||||
federation_sender_instances = config.get("federation_sender_instances") or []
|
federation_sender_instances = config.get("federation_sender_instances") or []
|
||||||
self.federation_shard_config = ShardedFederationSendingConfig(
|
self.federation_shard_config = ShardedWorkerHandlingConfig(
|
||||||
federation_sender_instances
|
federation_sender_instances
|
||||||
)
|
)
|
||||||
|
|
||||||
|
@ -14,7 +14,7 @@
|
|||||||
# See the License for the specific language governing permissions and
|
# See the License for the specific language governing permissions and
|
||||||
# limitations under the License.
|
# limitations under the License.
|
||||||
|
|
||||||
from ._base import Config
|
from ._base import Config, ShardedWorkerHandlingConfig
|
||||||
|
|
||||||
|
|
||||||
class PushConfig(Config):
|
class PushConfig(Config):
|
||||||
@ -24,6 +24,9 @@ class PushConfig(Config):
|
|||||||
push_config = config.get("push", {})
|
push_config = config.get("push", {})
|
||||||
self.push_include_content = push_config.get("include_content", True)
|
self.push_include_content = push_config.get("include_content", True)
|
||||||
|
|
||||||
|
pusher_instances = config.get("pusher_instances") or []
|
||||||
|
self.pusher_shard_config = ShardedWorkerHandlingConfig(pusher_instances)
|
||||||
|
|
||||||
# There was a a 'redact_content' setting but mistakenly read from the
|
# There was a a 'redact_content' setting but mistakenly read from the
|
||||||
# 'email'section'. Check for the flag in the 'push' section, and log,
|
# 'email'section'. Check for the flag in the 'push' section, and log,
|
||||||
# but do not honour it to avoid nasty surprises when people upgrade.
|
# but do not honour it to avoid nasty surprises when people upgrade.
|
||||||
|
@ -197,7 +197,7 @@ class FederationSender(object):
|
|||||||
destinations = {
|
destinations = {
|
||||||
d
|
d
|
||||||
for d in destinations
|
for d in destinations
|
||||||
if self._federation_shard_config.should_send_to(
|
if self._federation_shard_config.should_handle(
|
||||||
self._instance_name, d
|
self._instance_name, d
|
||||||
)
|
)
|
||||||
}
|
}
|
||||||
@ -335,7 +335,7 @@ class FederationSender(object):
|
|||||||
d
|
d
|
||||||
for d in domains
|
for d in domains
|
||||||
if d != self.server_name
|
if d != self.server_name
|
||||||
and self._federation_shard_config.should_send_to(self._instance_name, d)
|
and self._federation_shard_config.should_handle(self._instance_name, d)
|
||||||
]
|
]
|
||||||
if not domains:
|
if not domains:
|
||||||
return
|
return
|
||||||
@ -441,7 +441,7 @@ class FederationSender(object):
|
|||||||
for destination in destinations:
|
for destination in destinations:
|
||||||
if destination == self.server_name:
|
if destination == self.server_name:
|
||||||
continue
|
continue
|
||||||
if not self._federation_shard_config.should_send_to(
|
if not self._federation_shard_config.should_handle(
|
||||||
self._instance_name, destination
|
self._instance_name, destination
|
||||||
):
|
):
|
||||||
continue
|
continue
|
||||||
@ -460,7 +460,7 @@ class FederationSender(object):
|
|||||||
if destination == self.server_name:
|
if destination == self.server_name:
|
||||||
continue
|
continue
|
||||||
|
|
||||||
if not self._federation_shard_config.should_send_to(
|
if not self._federation_shard_config.should_handle(
|
||||||
self._instance_name, destination
|
self._instance_name, destination
|
||||||
):
|
):
|
||||||
continue
|
continue
|
||||||
@ -486,7 +486,7 @@ class FederationSender(object):
|
|||||||
logger.info("Not sending EDU to ourselves")
|
logger.info("Not sending EDU to ourselves")
|
||||||
return
|
return
|
||||||
|
|
||||||
if not self._federation_shard_config.should_send_to(
|
if not self._federation_shard_config.should_handle(
|
||||||
self._instance_name, destination
|
self._instance_name, destination
|
||||||
):
|
):
|
||||||
return
|
return
|
||||||
@ -507,7 +507,7 @@ class FederationSender(object):
|
|||||||
edu: edu to send
|
edu: edu to send
|
||||||
key: clobbering key for this edu
|
key: clobbering key for this edu
|
||||||
"""
|
"""
|
||||||
if not self._federation_shard_config.should_send_to(
|
if not self._federation_shard_config.should_handle(
|
||||||
self._instance_name, edu.destination
|
self._instance_name, edu.destination
|
||||||
):
|
):
|
||||||
return
|
return
|
||||||
@ -523,7 +523,7 @@ class FederationSender(object):
|
|||||||
logger.warning("Not sending device update to ourselves")
|
logger.warning("Not sending device update to ourselves")
|
||||||
return
|
return
|
||||||
|
|
||||||
if not self._federation_shard_config.should_send_to(
|
if not self._federation_shard_config.should_handle(
|
||||||
self._instance_name, destination
|
self._instance_name, destination
|
||||||
):
|
):
|
||||||
return
|
return
|
||||||
@ -541,7 +541,7 @@ class FederationSender(object):
|
|||||||
logger.warning("Not waking up ourselves")
|
logger.warning("Not waking up ourselves")
|
||||||
return
|
return
|
||||||
|
|
||||||
if not self._federation_shard_config.should_send_to(
|
if not self._federation_shard_config.should_handle(
|
||||||
self._instance_name, destination
|
self._instance_name, destination
|
||||||
):
|
):
|
||||||
return
|
return
|
||||||
|
@ -78,7 +78,7 @@ class PerDestinationQueue(object):
|
|||||||
self._federation_shard_config = hs.config.federation.federation_shard_config
|
self._federation_shard_config = hs.config.federation.federation_shard_config
|
||||||
|
|
||||||
self._should_send_on_this_instance = True
|
self._should_send_on_this_instance = True
|
||||||
if not self._federation_shard_config.should_send_to(
|
if not self._federation_shard_config.should_handle(
|
||||||
self._instance_name, destination
|
self._instance_name, destination
|
||||||
):
|
):
|
||||||
# We don't raise an exception here to avoid taking out any other
|
# We don't raise an exception here to avoid taking out any other
|
||||||
|
@ -15,13 +15,12 @@
|
|||||||
# limitations under the License.
|
# limitations under the License.
|
||||||
|
|
||||||
import logging
|
import logging
|
||||||
from collections import defaultdict
|
from typing import TYPE_CHECKING, Dict, Union
|
||||||
from threading import Lock
|
|
||||||
from typing import Dict, Tuple, Union
|
from prometheus_client import Gauge
|
||||||
|
|
||||||
from twisted.internet import defer
|
from twisted.internet import defer
|
||||||
|
|
||||||
from synapse.metrics import LaterGauge
|
|
||||||
from synapse.metrics.background_process_metrics import run_as_background_process
|
from synapse.metrics.background_process_metrics import run_as_background_process
|
||||||
from synapse.push import PusherConfigException
|
from synapse.push import PusherConfigException
|
||||||
from synapse.push.emailpusher import EmailPusher
|
from synapse.push.emailpusher import EmailPusher
|
||||||
@ -29,9 +28,18 @@ from synapse.push.httppusher import HttpPusher
|
|||||||
from synapse.push.pusher import PusherFactory
|
from synapse.push.pusher import PusherFactory
|
||||||
from synapse.util.async_helpers import concurrently_execute
|
from synapse.util.async_helpers import concurrently_execute
|
||||||
|
|
||||||
|
if TYPE_CHECKING:
|
||||||
|
from synapse.server import HomeServer
|
||||||
|
|
||||||
|
|
||||||
logger = logging.getLogger(__name__)
|
logger = logging.getLogger(__name__)
|
||||||
|
|
||||||
|
|
||||||
|
synapse_pushers = Gauge(
|
||||||
|
"synapse_pushers", "Number of active synapse pushers", ["kind", "app_id"]
|
||||||
|
)
|
||||||
|
|
||||||
|
|
||||||
class PusherPool:
|
class PusherPool:
|
||||||
"""
|
"""
|
||||||
The pusher pool. This is responsible for dispatching notifications of new events to
|
The pusher pool. This is responsible for dispatching notifications of new events to
|
||||||
@ -47,36 +55,20 @@ class PusherPool:
|
|||||||
Pusher.on_new_receipts are not expected to return deferreds.
|
Pusher.on_new_receipts are not expected to return deferreds.
|
||||||
"""
|
"""
|
||||||
|
|
||||||
def __init__(self, _hs):
|
def __init__(self, hs: "HomeServer"):
|
||||||
self.hs = _hs
|
self.hs = hs
|
||||||
self.pusher_factory = PusherFactory(_hs)
|
self.pusher_factory = PusherFactory(hs)
|
||||||
self._should_start_pushers = _hs.config.start_pushers
|
self._should_start_pushers = hs.config.start_pushers
|
||||||
self.store = self.hs.get_datastore()
|
self.store = self.hs.get_datastore()
|
||||||
self.clock = self.hs.get_clock()
|
self.clock = self.hs.get_clock()
|
||||||
|
|
||||||
|
# We shard the handling of push notifications by user ID.
|
||||||
|
self._pusher_shard_config = hs.config.push.pusher_shard_config
|
||||||
|
self._instance_name = hs.get_instance_name()
|
||||||
|
|
||||||
# map from user id to app_id:pushkey to pusher
|
# map from user id to app_id:pushkey to pusher
|
||||||
self.pushers = {} # type: Dict[str, Dict[str, Union[HttpPusher, EmailPusher]]]
|
self.pushers = {} # type: Dict[str, Dict[str, Union[HttpPusher, EmailPusher]]]
|
||||||
|
|
||||||
# a lock for the pushers dict, since `count_pushers` is called from an different
|
|
||||||
# and we otherwise get concurrent modification errors
|
|
||||||
self._pushers_lock = Lock()
|
|
||||||
|
|
||||||
def count_pushers():
|
|
||||||
results = defaultdict(int) # type: Dict[Tuple[str, str], int]
|
|
||||||
with self._pushers_lock:
|
|
||||||
for pushers in self.pushers.values():
|
|
||||||
for pusher in pushers.values():
|
|
||||||
k = (type(pusher).__name__, pusher.app_id)
|
|
||||||
results[k] += 1
|
|
||||||
return results
|
|
||||||
|
|
||||||
LaterGauge(
|
|
||||||
name="synapse_pushers",
|
|
||||||
desc="the number of active pushers",
|
|
||||||
labels=["kind", "app_id"],
|
|
||||||
caller=count_pushers,
|
|
||||||
)
|
|
||||||
|
|
||||||
def start(self):
|
def start(self):
|
||||||
"""Starts the pushers off in a background process.
|
"""Starts the pushers off in a background process.
|
||||||
"""
|
"""
|
||||||
@ -104,6 +96,7 @@ class PusherPool:
|
|||||||
Returns:
|
Returns:
|
||||||
Deferred[EmailPusher|HttpPusher]
|
Deferred[EmailPusher|HttpPusher]
|
||||||
"""
|
"""
|
||||||
|
|
||||||
time_now_msec = self.clock.time_msec()
|
time_now_msec = self.clock.time_msec()
|
||||||
|
|
||||||
# we try to create the pusher just to validate the config: it
|
# we try to create the pusher just to validate the config: it
|
||||||
@ -176,6 +169,9 @@ class PusherPool:
|
|||||||
access_tokens (Iterable[int]): access token *ids* to remove pushers
|
access_tokens (Iterable[int]): access token *ids* to remove pushers
|
||||||
for
|
for
|
||||||
"""
|
"""
|
||||||
|
if not self._pusher_shard_config.should_handle(self._instance_name, user_id):
|
||||||
|
return
|
||||||
|
|
||||||
tokens = set(access_tokens)
|
tokens = set(access_tokens)
|
||||||
for p in (yield self.store.get_pushers_by_user_id(user_id)):
|
for p in (yield self.store.get_pushers_by_user_id(user_id)):
|
||||||
if p["access_token"] in tokens:
|
if p["access_token"] in tokens:
|
||||||
@ -237,6 +233,9 @@ class PusherPool:
|
|||||||
if not self._should_start_pushers:
|
if not self._should_start_pushers:
|
||||||
return
|
return
|
||||||
|
|
||||||
|
if not self._pusher_shard_config.should_handle(self._instance_name, user_id):
|
||||||
|
return
|
||||||
|
|
||||||
resultlist = yield self.store.get_pushers_by_app_id_and_pushkey(app_id, pushkey)
|
resultlist = yield self.store.get_pushers_by_app_id_and_pushkey(app_id, pushkey)
|
||||||
|
|
||||||
pusher_dict = None
|
pusher_dict = None
|
||||||
@ -275,6 +274,11 @@ class PusherPool:
|
|||||||
Returns:
|
Returns:
|
||||||
Deferred[EmailPusher|HttpPusher]
|
Deferred[EmailPusher|HttpPusher]
|
||||||
"""
|
"""
|
||||||
|
if not self._pusher_shard_config.should_handle(
|
||||||
|
self._instance_name, pusherdict["user_name"]
|
||||||
|
):
|
||||||
|
return
|
||||||
|
|
||||||
try:
|
try:
|
||||||
p = self.pusher_factory.create_pusher(pusherdict)
|
p = self.pusher_factory.create_pusher(pusherdict)
|
||||||
except PusherConfigException as e:
|
except PusherConfigException as e:
|
||||||
@ -298,11 +302,12 @@ class PusherPool:
|
|||||||
|
|
||||||
appid_pushkey = "%s:%s" % (pusherdict["app_id"], pusherdict["pushkey"])
|
appid_pushkey = "%s:%s" % (pusherdict["app_id"], pusherdict["pushkey"])
|
||||||
|
|
||||||
with self._pushers_lock:
|
byuser = self.pushers.setdefault(pusherdict["user_name"], {})
|
||||||
byuser = self.pushers.setdefault(pusherdict["user_name"], {})
|
if appid_pushkey in byuser:
|
||||||
if appid_pushkey in byuser:
|
byuser[appid_pushkey].on_stop()
|
||||||
byuser[appid_pushkey].on_stop()
|
byuser[appid_pushkey] = p
|
||||||
byuser[appid_pushkey] = p
|
|
||||||
|
synapse_pushers.labels(type(p).__name__, p.app_id).inc()
|
||||||
|
|
||||||
# Check if there *may* be push to process. We do this as this check is a
|
# Check if there *may* be push to process. We do this as this check is a
|
||||||
# lot cheaper to do than actually fetching the exact rows we need to
|
# lot cheaper to do than actually fetching the exact rows we need to
|
||||||
@ -330,9 +335,10 @@ class PusherPool:
|
|||||||
|
|
||||||
if appid_pushkey in byuser:
|
if appid_pushkey in byuser:
|
||||||
logger.info("Stopping pusher %s / %s", user_id, appid_pushkey)
|
logger.info("Stopping pusher %s / %s", user_id, appid_pushkey)
|
||||||
byuser[appid_pushkey].on_stop()
|
pusher = byuser.pop(appid_pushkey)
|
||||||
with self._pushers_lock:
|
pusher.on_stop()
|
||||||
del byuser[appid_pushkey]
|
|
||||||
|
synapse_pushers.labels(type(pusher).__name__, pusher.app_id).dec()
|
||||||
|
|
||||||
yield self.store.delete_pusher_by_app_id_pushkey_user_id(
|
yield self.store.delete_pusher_by_app_id_pushkey_user_id(
|
||||||
app_id, pushkey, user_id
|
app_id, pushkey, user_id
|
||||||
|
193
tests/replication/test_pusher_shard.py
Normal file
193
tests/replication/test_pusher_shard.py
Normal file
@ -0,0 +1,193 @@
|
|||||||
|
# -*- coding: utf-8 -*-
|
||||||
|
# Copyright 2020 The Matrix.org Foundation C.I.C.
|
||||||
|
#
|
||||||
|
# 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 mock import Mock
|
||||||
|
|
||||||
|
from twisted.internet import defer
|
||||||
|
|
||||||
|
from synapse.rest import admin
|
||||||
|
from synapse.rest.client.v1 import login, room
|
||||||
|
|
||||||
|
from tests.replication._base import BaseMultiWorkerStreamTestCase
|
||||||
|
|
||||||
|
logger = logging.getLogger(__name__)
|
||||||
|
|
||||||
|
|
||||||
|
class PusherShardTestCase(BaseMultiWorkerStreamTestCase):
|
||||||
|
"""Checks pusher sharding works
|
||||||
|
"""
|
||||||
|
|
||||||
|
servlets = [
|
||||||
|
admin.register_servlets_for_client_rest_resource,
|
||||||
|
room.register_servlets,
|
||||||
|
login.register_servlets,
|
||||||
|
]
|
||||||
|
|
||||||
|
def prepare(self, reactor, clock, hs):
|
||||||
|
# Register a user who sends a message that we'll get notified about
|
||||||
|
self.other_user_id = self.register_user("otheruser", "pass")
|
||||||
|
self.other_access_token = self.login("otheruser", "pass")
|
||||||
|
|
||||||
|
def default_config(self):
|
||||||
|
conf = super().default_config()
|
||||||
|
conf["start_pushers"] = False
|
||||||
|
return conf
|
||||||
|
|
||||||
|
def _create_pusher_and_send_msg(self, localpart):
|
||||||
|
# Create a user that will get push notifications
|
||||||
|
user_id = self.register_user(localpart, "pass")
|
||||||
|
access_token = self.login(localpart, "pass")
|
||||||
|
|
||||||
|
# Register a pusher
|
||||||
|
user_dict = self.get_success(
|
||||||
|
self.hs.get_datastore().get_user_by_access_token(access_token)
|
||||||
|
)
|
||||||
|
token_id = user_dict["token_id"]
|
||||||
|
|
||||||
|
self.get_success(
|
||||||
|
self.hs.get_pusherpool().add_pusher(
|
||||||
|
user_id=user_id,
|
||||||
|
access_token=token_id,
|
||||||
|
kind="http",
|
||||||
|
app_id="m.http",
|
||||||
|
app_display_name="HTTP Push Notifications",
|
||||||
|
device_display_name="pushy push",
|
||||||
|
pushkey="a@example.com",
|
||||||
|
lang=None,
|
||||||
|
data={"url": "https://push.example.com/push"},
|
||||||
|
)
|
||||||
|
)
|
||||||
|
|
||||||
|
self.pump()
|
||||||
|
|
||||||
|
# Create a room
|
||||||
|
room = self.helper.create_room_as(user_id, tok=access_token)
|
||||||
|
|
||||||
|
# The other user joins
|
||||||
|
self.helper.join(
|
||||||
|
room=room, user=self.other_user_id, tok=self.other_access_token
|
||||||
|
)
|
||||||
|
|
||||||
|
# The other user sends some messages
|
||||||
|
response = self.helper.send(room, body="Hi!", tok=self.other_access_token)
|
||||||
|
event_id = response["event_id"]
|
||||||
|
|
||||||
|
return event_id
|
||||||
|
|
||||||
|
def test_send_push_single_worker(self):
|
||||||
|
"""Test that registration works when using a pusher worker.
|
||||||
|
"""
|
||||||
|
http_client_mock = Mock(spec_set=["post_json_get_json"])
|
||||||
|
http_client_mock.post_json_get_json.side_effect = lambda *_, **__: defer.succeed(
|
||||||
|
{}
|
||||||
|
)
|
||||||
|
|
||||||
|
self.make_worker_hs(
|
||||||
|
"synapse.app.pusher",
|
||||||
|
{"start_pushers": True},
|
||||||
|
proxied_http_client=http_client_mock,
|
||||||
|
)
|
||||||
|
|
||||||
|
event_id = self._create_pusher_and_send_msg("user")
|
||||||
|
|
||||||
|
# Advance time a bit, so the pusher will register something has happened
|
||||||
|
self.pump()
|
||||||
|
|
||||||
|
http_client_mock.post_json_get_json.assert_called_once()
|
||||||
|
self.assertEqual(
|
||||||
|
http_client_mock.post_json_get_json.call_args[0][0],
|
||||||
|
"https://push.example.com/push",
|
||||||
|
)
|
||||||
|
self.assertEqual(
|
||||||
|
event_id,
|
||||||
|
http_client_mock.post_json_get_json.call_args[0][1]["notification"][
|
||||||
|
"event_id"
|
||||||
|
],
|
||||||
|
)
|
||||||
|
|
||||||
|
def test_send_push_multiple_workers(self):
|
||||||
|
"""Test that registration works when using sharded pusher workers.
|
||||||
|
"""
|
||||||
|
http_client_mock1 = Mock(spec_set=["post_json_get_json"])
|
||||||
|
http_client_mock1.post_json_get_json.side_effect = lambda *_, **__: defer.succeed(
|
||||||
|
{}
|
||||||
|
)
|
||||||
|
|
||||||
|
self.make_worker_hs(
|
||||||
|
"synapse.app.pusher",
|
||||||
|
{
|
||||||
|
"start_pushers": True,
|
||||||
|
"worker_name": "pusher1",
|
||||||
|
"pusher_instances": ["pusher1", "pusher2"],
|
||||||
|
},
|
||||||
|
proxied_http_client=http_client_mock1,
|
||||||
|
)
|
||||||
|
|
||||||
|
http_client_mock2 = Mock(spec_set=["post_json_get_json"])
|
||||||
|
http_client_mock2.post_json_get_json.side_effect = lambda *_, **__: defer.succeed(
|
||||||
|
{}
|
||||||
|
)
|
||||||
|
|
||||||
|
self.make_worker_hs(
|
||||||
|
"synapse.app.pusher",
|
||||||
|
{
|
||||||
|
"start_pushers": True,
|
||||||
|
"worker_name": "pusher2",
|
||||||
|
"pusher_instances": ["pusher1", "pusher2"],
|
||||||
|
},
|
||||||
|
proxied_http_client=http_client_mock2,
|
||||||
|
)
|
||||||
|
|
||||||
|
# We choose a user name that we know should go to pusher1.
|
||||||
|
event_id = self._create_pusher_and_send_msg("user2")
|
||||||
|
|
||||||
|
# Advance time a bit, so the pusher will register something has happened
|
||||||
|
self.pump()
|
||||||
|
|
||||||
|
http_client_mock1.post_json_get_json.assert_called_once()
|
||||||
|
http_client_mock2.post_json_get_json.assert_not_called()
|
||||||
|
self.assertEqual(
|
||||||
|
http_client_mock1.post_json_get_json.call_args[0][0],
|
||||||
|
"https://push.example.com/push",
|
||||||
|
)
|
||||||
|
self.assertEqual(
|
||||||
|
event_id,
|
||||||
|
http_client_mock1.post_json_get_json.call_args[0][1]["notification"][
|
||||||
|
"event_id"
|
||||||
|
],
|
||||||
|
)
|
||||||
|
|
||||||
|
http_client_mock1.post_json_get_json.reset_mock()
|
||||||
|
http_client_mock2.post_json_get_json.reset_mock()
|
||||||
|
|
||||||
|
# Now we choose a user name that we know should go to pusher2.
|
||||||
|
event_id = self._create_pusher_and_send_msg("user4")
|
||||||
|
|
||||||
|
# Advance time a bit, so the pusher will register something has happened
|
||||||
|
self.pump()
|
||||||
|
|
||||||
|
http_client_mock1.post_json_get_json.assert_not_called()
|
||||||
|
http_client_mock2.post_json_get_json.assert_called_once()
|
||||||
|
self.assertEqual(
|
||||||
|
http_client_mock2.post_json_get_json.call_args[0][0],
|
||||||
|
"https://push.example.com/push",
|
||||||
|
)
|
||||||
|
self.assertEqual(
|
||||||
|
event_id,
|
||||||
|
http_client_mock2.post_json_get_json.call_args[0][1]["notification"][
|
||||||
|
"event_id"
|
||||||
|
],
|
||||||
|
)
|
Loading…
Reference in New Issue
Block a user