Matrix-Mirrors/anonymousland-synapse
1
0
Fork 0
mirror of https://git.anonymousland.org/anonymousland/synapse.git synced 2024-10-01 11:49:51 -04:00
Code Issues Packages Projects Releases Wiki Activity
1eb8dcf4c9
anonymousland-synapse/docker/configure_workers_and_start.py

656 lines
25 KiB
Python
Raw Normal View History

Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
#!/usr/bin/env python
# Copyright 2021 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.
# This script reads environment variables and generates a shared Synapse worker,
# nginx and supervisord configs depending on the workers requested.
#
# The environment variables it reads are:
# * SYNAPSE_SERVER_NAME: The desired server_name of the homeserver.
# * SYNAPSE_REPORT_STATS: Whether to report stats.
# * SYNAPSE_WORKER_TYPES: A comma separated list of worker names as specified in WORKER_CONFIG
# below. Leave empty for no workers, or set to '*' for all possible workers.
Support registering Application Services when running with workers under Complement. (#12826) Co-authored-by: Patrick Cloke <clokep@users.noreply.github.com>
2022-05-23 09:11:06 -04:00
# * SYNAPSE_AS_REGISTRATION_DIR: If specified, a directory in which .yaml and .yml files
# will be treated as Application Service registration files.
Remove Caddy from the Synapse workers image used in Complement. (#12818)
2022-05-23 05:29:24 -04:00
# * SYNAPSE_TLS_CERT: Path to a TLS certificate in PEM format.
# * SYNAPSE_TLS_KEY: Path to a TLS key. If this and SYNAPSE_TLS_CERT are specified,
# Nginx will be configured to serve TLS on port 8448.
Improve startup times in Complement test runs against workers, particularly in CPU-constrained environments. (#13127) Co-authored-by: Richard van der Hoff <1389908+richvdh@users.noreply.github.com>
2022-06-30 07:58:12 -04:00
# * SYNAPSE_USE_EXPERIMENTAL_FORKING_LAUNCHER: Whether to use the forking launcher,
# only intended for usage in Complement at the moment.
# No stability guarantees are provided.
Add the ability to set the log level using the `SYNAPSE_TEST_LOG_LEVEL` environment when using `complement.sh`. (#13152)
2022-07-05 05:46:20 -04:00
# * SYNAPSE_LOG_LEVEL: Set this to DEBUG, INFO, WARNING or ERROR to change the
# log level. INFO is the default.
# * SYNAPSE_LOG_SENSITIVE: If unset, SQL and SQL values won't be logged,
# regardless of the SYNAPSE_LOG_LEVEL setting.
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
#
# NOTE: According to Complement's ENTRYPOINT expectations for a homeserver image (as defined
# in the project's README), this script may be run multiple times, and functionality should
# continue to work if so.
import os
Set LD_PRELOAD to load jemalloc in Dockerfile-workers. (#14182)
2022-10-14 08:29:49 -04:00
import platform
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
import subprocess
import sys
Support registering Application Services when running with workers under Complement. (#12826) Co-authored-by: Patrick Cloke <clokep@users.noreply.github.com>
2022-05-23 09:11:06 -04:00
from pathlib import Path
Add the ability to set the log level using the `SYNAPSE_TEST_LOG_LEVEL` environment when using `complement.sh`. (#13152)
2022-07-05 05:46:20 -04:00
from typing import Any, Dict, List, Mapping, MutableMapping, NoReturn, Optional, Set
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
import yaml
Merge the Complement testing Docker images into a single, multi-purpose image. (#12881) Co-authored-by: Richard van der Hoff <1389908+richvdh@users.noreply.github.com>
2022-06-08 05:57:05 -04:00
from jinja2 import Environment, FileSystemLoader
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
MAIN_PROCESS_HTTP_LISTENER_PORT = 8080
Add missing type definitions for scripts in docker folder (#12280) Signed-off-by: Jorge Florian <jafn28@gmail.com>
2022-04-08 06:10:58 -04:00
WORKERS_CONFIG: Dict[str, Dict[str, Any]] = {
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
"pusher": {
"app": "synapse.app.pusher",
"listener_resources": [],
"endpoint_patterns": [],
"shared_extra_conf": {"start_pushers": False},
"worker_extra_conf": "",
},
"user_dir": {
Use updated `update_user_directory_from_worker` options in Complement tests. (#13069)
2022-06-15 10:54:32 -04:00
"app": "synapse.app.generic_worker",
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
"listener_resources": ["client"],
"endpoint_patterns": [
Add support for `/_matrix/media/v3` APIs (#11371) * Add support for `/_matrix/media/v3` APIs Signed-off-by: Aaron Raimist <aaron@raim.ist> * Update `workers.md` to use v3 client and media APIs Signed-off-by: Aaron Raimist <aaron@raim.ist> * Add changelog Signed-off-by: Aaron Raimist <aaron@raim.ist>
2021-11-17 10:30:24 -05:00
"^/_matrix/client/(api/v1|r0|v3|unstable)/user_directory/search$"
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
],
Use updated `update_user_directory_from_worker` options in Complement tests. (#13069)
2022-06-15 10:54:32 -04:00
"shared_extra_conf": {"update_user_directory_from_worker": "user_dir1"},
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
"worker_extra_conf": "",
},
"media_repository": {
"app": "synapse.app.media_repository",
"listener_resources": ["media"],
"endpoint_patterns": [
"^/_matrix/media/",
"^/_synapse/admin/v1/purge_media_cache$",
"^/_synapse/admin/v1/room/.*/media.*$",
"^/_synapse/admin/v1/user/.*/media.*$",
"^/_synapse/admin/v1/media/.*$",
"^/_synapse/admin/v1/quarantine_media/.*$",
],
"shared_extra_conf": {"enable_media_repo": False},
"worker_extra_conf": "enable_media_repo: true",
},
"appservice": {
Add the `notify_appservices_from_worker` configuration option (superseding `notify_appservices`) to allow a generic worker to be designated as the worker to send traffic to Application Services. (#12452)
2022-05-06 06:43:53 -04:00
"app": "synapse.app.generic_worker",
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
"listener_resources": [],
"endpoint_patterns": [],
Use updated `update_user_directory_from_worker` options in Complement tests. (#13069)
2022-06-15 10:54:32 -04:00
"shared_extra_conf": {"notify_appservices_from_worker": "appservice1"},
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
"worker_extra_conf": "",
},
"federation_sender": {
"app": "synapse.app.federation_sender",
"listener_resources": [],
"endpoint_patterns": [],
"shared_extra_conf": {"send_federation": False},
"worker_extra_conf": "",
},
"synchrotron": {
"app": "synapse.app.generic_worker",
"listener_resources": ["client"],
"endpoint_patterns": [
Add support for `/_matrix/media/v3` APIs (#11371) * Add support for `/_matrix/media/v3` APIs Signed-off-by: Aaron Raimist <aaron@raim.ist> * Update `workers.md` to use v3 client and media APIs Signed-off-by: Aaron Raimist <aaron@raim.ist> * Add changelog Signed-off-by: Aaron Raimist <aaron@raim.ist>
2021-11-17 10:30:24 -05:00
"^/_matrix/client/(v2_alpha|r0|v3)/sync$",
"^/_matrix/client/(api/v1|v2_alpha|r0|v3)/events$",
"^/_matrix/client/(api/v1|r0|v3)/initialSync$",
"^/_matrix/client/(api/v1|r0|v3)/rooms/[^/]+/initialSync$",
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
],
"shared_extra_conf": {},
"worker_extra_conf": "",
},
Mark /relations endpoint as usable on workers. (#14028) Co-authored-by: Eric Eastwood <erice@element.io>
2022-10-12 06:46:13 -04:00
"client_reader": {
"app": "synapse.app.generic_worker",
"listener_resources": ["client"],
"endpoint_patterns": [
"^/_matrix/client/(api/v1|r0|v3|unstable)/publicRooms$",
"^/_matrix/client/(api/v1|r0|v3|unstable)/rooms/.*/joined_members$",
"^/_matrix/client/(api/v1|r0|v3|unstable)/rooms/.*/context/.*$",
"^/_matrix/client/(api/v1|r0|v3|unstable)/rooms/.*/members$",
"^/_matrix/client/(api/v1|r0|v3|unstable)/rooms/.*/state$",
"^/_matrix/client/v1/rooms/.*/hierarchy$",
"^/_matrix/client/(v1|unstable)/rooms/.*/relations/",
Stabilize the threads API. (#14175) Stabilize the threads API (MSC3856) by supporting (only) the v1 path for the endpoint. This also marks the API as safe for workers since it is a read-only API.
2022-10-14 07:16:50 -04:00
"^/_matrix/client/v1/rooms/.*/threads$",
Mark /relations endpoint as usable on workers. (#14028) Co-authored-by: Eric Eastwood <erice@element.io>
2022-10-12 06:46:13 -04:00
"^/_matrix/client/(api/v1|r0|v3|unstable)/login$",
"^/_matrix/client/(api/v1|r0|v3|unstable)/account/3pid$",
"^/_matrix/client/(api/v1|r0|v3|unstable)/account/whoami$",
"^/_matrix/client/versions$",
"^/_matrix/client/(api/v1|r0|v3|unstable)/voip/turnServer$",
"^/_matrix/client/(r0|v3|unstable)/register$",
"^/_matrix/client/(r0|v3|unstable)/auth/.*/fallback/web$",
"^/_matrix/client/(api/v1|r0|v3|unstable)/rooms/.*/messages$",
"^/_matrix/client/(api/v1|r0|v3|unstable)/rooms/.*/event",
"^/_matrix/client/(api/v1|r0|v3|unstable)/joined_rooms",
"^/_matrix/client/(api/v1|r0|v3|unstable/.*)/rooms/.*/aliases",
"^/_matrix/client/(api/v1|r0|v3|unstable)/search",
],
"shared_extra_conf": {},
"worker_extra_conf": "",
},
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
"federation_reader": {
"app": "synapse.app.generic_worker",
"listener_resources": ["federation"],
"endpoint_patterns": [
"^/_matrix/federation/(v1|v2)/event/",
"^/_matrix/federation/(v1|v2)/state/",
"^/_matrix/federation/(v1|v2)/state_ids/",
"^/_matrix/federation/(v1|v2)/backfill/",
"^/_matrix/federation/(v1|v2)/get_missing_events/",
"^/_matrix/federation/(v1|v2)/publicRooms",
"^/_matrix/federation/(v1|v2)/query/",
"^/_matrix/federation/(v1|v2)/make_join/",
"^/_matrix/federation/(v1|v2)/make_leave/",
"^/_matrix/federation/(v1|v2)/send_join/",
"^/_matrix/federation/(v1|v2)/send_leave/",
"^/_matrix/federation/(v1|v2)/invite/",
"^/_matrix/federation/(v1|v2)/query_auth/",
"^/_matrix/federation/(v1|v2)/event_auth/",
"^/_matrix/federation/(v1|v2)/exchange_third_party_invite/",
"^/_matrix/federation/(v1|v2)/user/devices/",
"^/_matrix/federation/(v1|v2)/get_groups_publicised$",
"^/_matrix/key/v2/query",
],
"shared_extra_conf": {},
"worker_extra_conf": "",
},
"federation_inbound": {
"app": "synapse.app.generic_worker",
"listener_resources": ["federation"],
"endpoint_patterns": ["/_matrix/federation/(v1|v2)/send/"],
"shared_extra_conf": {},
"worker_extra_conf": "",
},
"event_persister": {
"app": "synapse.app.generic_worker",
"listener_resources": ["replication"],
"endpoint_patterns": [],
"shared_extra_conf": {},
"worker_extra_conf": "",
},
"background_worker": {
"app": "synapse.app.generic_worker",
"listener_resources": [],
"endpoint_patterns": [],
# This worker cannot be sharded. Therefore there should only ever be one background
# worker, and it should be named background_worker1
"shared_extra_conf": {"run_background_tasks_on": "background_worker1"},
"worker_extra_conf": "",
},
"event_creator": {
"app": "synapse.app.generic_worker",
"listener_resources": ["client"],
"endpoint_patterns": [
Add support for `/_matrix/media/v3` APIs (#11371) * Add support for `/_matrix/media/v3` APIs Signed-off-by: Aaron Raimist <aaron@raim.ist> * Update `workers.md` to use v3 client and media APIs Signed-off-by: Aaron Raimist <aaron@raim.ist> * Add changelog Signed-off-by: Aaron Raimist <aaron@raim.ist>
2021-11-17 10:30:24 -05:00
"^/_matrix/client/(api/v1|r0|v3|unstable)/rooms/.*/redact",
"^/_matrix/client/(api/v1|r0|v3|unstable)/rooms/.*/send",
"^/_matrix/client/(api/v1|r0|v3|unstable)/rooms/.*/(join|invite|leave|ban|unban|kick)$",
"^/_matrix/client/(api/v1|r0|v3|unstable)/join/",
"^/_matrix/client/(api/v1|r0|v3|unstable)/profile/",
Add the `batch_send` endpoint to generic workers (#12868)
2022-05-25 05:51:07 -04:00
"^/_matrix/client/(v1|unstable/org.matrix.msc2716)/rooms/.*/batch_send",
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
],
"shared_extra_conf": {},
"worker_extra_conf": "",
},
"frontend_proxy": {
"app": "synapse.app.frontend_proxy",
"listener_resources": ["client", "replication"],
Add support for `/_matrix/media/v3` APIs (#11371) * Add support for `/_matrix/media/v3` APIs Signed-off-by: Aaron Raimist <aaron@raim.ist> * Update `workers.md` to use v3 client and media APIs Signed-off-by: Aaron Raimist <aaron@raim.ist> * Add changelog Signed-off-by: Aaron Raimist <aaron@raim.ist>
2021-11-17 10:30:24 -05:00
"endpoint_patterns": ["^/_matrix/client/(api/v1|r0|v3|unstable)/keys/upload"],
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
"shared_extra_conf": {},
"worker_extra_conf": (
"worker_main_http_uri: http://127.0.0.1:%d"
Fix frontend_proxy jinja script in docker workers (#10783)
2021-09-08 12:24:53 -04:00
% (MAIN_PROCESS_HTTP_LISTENER_PORT,)
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
),
},
}
# Templates for sections that may be inserted multiple times in config files
NGINX_LOCATION_CONFIG_BLOCK = """
Run complement with Synapse workers manually. (#10039) Adds an option to complement.sh to run Synapse in worker mode (instead of the default monolith mode).
2021-05-24 15:32:45 -04:00
location ~* {endpoint} {{
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
proxy_pass {upstream};
proxy_set_header X-Forwarded-For $remote_addr;
proxy_set_header X-Forwarded-Proto $scheme;
proxy_set_header Host $host;
Run complement with Synapse workers manually. (#10039) Adds an option to complement.sh to run Synapse in worker mode (instead of the default monolith mode).
2021-05-24 15:32:45 -04:00
}}
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
"""
NGINX_UPSTREAM_CONFIG_BLOCK = """
Run complement with Synapse workers manually. (#10039) Adds an option to complement.sh to run Synapse in worker mode (instead of the default monolith mode).
2021-05-24 15:32:45 -04:00
upstream {upstream_worker_type} {{
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
{body}
Run complement with Synapse workers manually. (#10039) Adds an option to complement.sh to run Synapse in worker mode (instead of the default monolith mode).
2021-05-24 15:32:45 -04:00
}}
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
"""
# Utility functions
disallow-untyped-defs in `docker` and `stubs` directories (#12528)
2022-04-25 08:32:35 -04:00
def log(txt: str) -> None:
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
"""Log something to the stdout.
Args:
txt: The text to log.
"""
print(txt)
disallow-untyped-defs in `docker` and `stubs` directories (#12528)
2022-04-25 08:32:35 -04:00
def error(txt: str) -> NoReturn:
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
"""Log something and exit with an error code.
Args:
txt: The text to log in error.
"""
log(txt)
sys.exit(2)
disallow-untyped-defs in `docker` and `stubs` directories (#12528)
2022-04-25 08:32:35 -04:00
def convert(src: str, dst: str, **template_vars: object) -> None:
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
"""Generate a file from a template
Args:
src: Path to the input file.
dst: Path to write to.
template_vars: The arguments to replace placeholder variables in the template with.
"""
# Read the template file
Merge the Complement testing Docker images into a single, multi-purpose image. (#12881) Co-authored-by: Richard van der Hoff <1389908+richvdh@users.noreply.github.com>
2022-06-08 05:57:05 -04:00
# We disable autoescape to prevent template variables from being escaped,
# as we're not using HTML.
env = Environment(loader=FileSystemLoader(os.path.dirname(src)), autoescape=False)
template = env.get_template(os.path.basename(src))
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
Merge the Complement testing Docker images into a single, multi-purpose image. (#12881) Co-authored-by: Richard van der Hoff <1389908+richvdh@users.noreply.github.com>
2022-06-08 05:57:05 -04:00
# Generate a string from the template.
rendered = template.render(**template_vars)
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
# Write the generated contents to a file
#
# We use append mode in case the files have already been written to by something else
# (for instance, as part of the instructions in a dockerfile).
with open(dst, "a") as outfile:
# In case the existing file doesn't end with a newline
outfile.write("\n")
outfile.write(rendered)
def add_sharding_to_shared_config(
shared_config: dict,
worker_type: str,
worker_name: str,
worker_port: int,
) -> None:
"""Given a dictionary representing a config file shared across all workers,
append sharded worker information to it for the current worker_type instance.
Args:
shared_config: The config dict that all worker instances share (after being converted to YAML)
worker_type: The type of worker (one of those defined in WORKERS_CONFIG).
worker_name: The name of the worker instance.
worker_port: The HTTP replication port that the worker instance is listening on.
"""
# The instance_map config field marks the workers that write to various replication streams
instance_map = shared_config.setdefault("instance_map", {})
# Worker-type specific sharding config
if worker_type == "pusher":
shared_config.setdefault("pusher_instances", []).append(worker_name)
elif worker_type == "federation_sender":
shared_config.setdefault("federation_sender_instances", []).append(worker_name)
elif worker_type == "event_persister":
# Event persisters write to the events stream, so we need to update
# the list of event stream writers
shared_config.setdefault("stream_writers", {}).setdefault("events", []).append(
worker_name
)
# Map of stream writer instance names to host/ports combos
instance_map[worker_name] = {
"host": "localhost",
"port": worker_port,
}
elif worker_type == "media_repository":
# The first configured media worker will run the media background jobs
shared_config.setdefault("media_instance_running_background_jobs", worker_name)
disallow-untyped-defs in `docker` and `stubs` directories (#12528)
2022-04-25 08:32:35 -04:00
def generate_base_homeserver_config() -> None:
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
"""Starts Synapse and generates a basic homeserver config, which will later be
modified for worker support.
Raises: CalledProcessError if calling start.py returned a non-zero exit code.
"""
# start.py already does this for us, so just call that.
# note that this script is copied in in the official, monolith dockerfile
os.environ["SYNAPSE_HTTP_PORT"] = str(MAIN_PROCESS_HTTP_LISTENER_PORT)
subprocess.check_output(["/usr/local/bin/python", "/start.py", "migrate_config"])
disallow-untyped-defs in `docker` and `stubs` directories (#12528)
2022-04-25 08:32:35 -04:00
def generate_worker_files(
environ: Mapping[str, str], config_path: str, data_dir: str
) -> None:
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
"""Read the desired list of workers from environment variables and generate
shared homeserver, nginx and supervisord configs.
Args:
disallow-untyped-defs in `docker` and `stubs` directories (#12528)
2022-04-25 08:32:35 -04:00
environ: os.environ instance.
Fix up healthcheck generation for workers docker image (#12405) This wasn't quite generating the right thing.
2022-04-11 09:38:58 -04:00
config_path: The location of the generated Synapse main worker config file.
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
data_dir: The location of the synapse data directory. Where log and
user-facing config files live.
"""
# Note that yaml cares about indentation, so care should be taken to insert lines
# into files at the correct indentation below.
# shared_config is the contents of a Synapse config file that will be shared amongst
# the main Synapse process as well as all workers.
# It is intended mainly for disabling functionality when certain workers are spun up,
# and adding a replication listener.
# First read the original config file and extract the listeners block. Then we'll add
Fix up healthcheck generation for workers docker image (#12405) This wasn't quite generating the right thing.
2022-04-11 09:38:58 -04:00
# another listener for replication. Later we'll write out the result to the shared
# config file.
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
listeners = [
{
"port": 9093,
"bind_address": "127.0.0.1",
"type": "http",
"resources": [{"names": ["replication"]}],
}
]
with open(config_path) as file_stream:
original_config = yaml.safe_load(file_stream)
original_listeners = original_config.get("listeners")
if original_listeners:
listeners += original_listeners
# The shared homeserver config. The contents of which will be inserted into the
# base shared worker jinja2 template.
#
# This config file will be passed to all workers, included Synapse's main process.
Dockerfile-workers: give the master its own log config (#12466) When we run a worker-mode synapse under docker, everything gets logged to stdout. Currently, output from the workers is tacked with a worker name, for example: ``` 2022-04-13 15:27:56,810 - worker:frontend_proxy1 - synapse.util.caches.lrucache - 154 - INFO - LruCache._expire_old_entries-0 - Dropped 0 items from caches ``` - note `worker:frontend_proxy1`. No such tag is applied to log lines from the master, which makes somewhat confusing reading. To fix this, we generate a dedicated log config file for the master in the same way that we do for the workers, and use that.
2022-04-13 15:50:08 -04:00
shared_config: Dict[str, Any] = {"listeners": listeners}
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
Refactor the Dockerfile-workers configuration script to use Jinja2 templates in Synapse workers' Supervisord blocks. (#13054) Co-authored-by: Richard van der Hoff <1389908+richvdh@users.noreply.github.com>
2022-06-27 06:43:20 -04:00
# List of dicts that describe workers.
# We pass this to the Supervisor template later to generate the appropriate
# program blocks.
worker_descriptors: List[Dict[str, Any]] = []
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
# Upstreams for load-balancing purposes. This dict takes the form of a worker type to the
# ports of each worker. For example:
# {
# worker_type: {1234, 1235, ...}}
# }
# and will be used to construct 'upstream' nginx directives.
Add missing type definitions for scripts in docker folder (#12280) Signed-off-by: Jorge Florian <jafn28@gmail.com>
2022-04-08 06:10:58 -04:00
nginx_upstreams: Dict[str, Set[int]] = {}
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
# A map of: {"endpoint": "upstream"}, where "upstream" is a str representing what will be
# placed after the proxy_pass directive. The main benefit to representing this data as a
# dict over a str is that we can easily deduplicate endpoints across multiple instances
# of the same worker.
#
# An nginx site config that will be amended to depending on the workers that are
# spun up. To be placed in /etc/nginx/conf.d.
nginx_locations = {}
# Read the desired worker configuration from the environment
Merge the Complement testing Docker images into a single, multi-purpose image. (#12881) Co-authored-by: Richard van der Hoff <1389908+richvdh@users.noreply.github.com>
2022-06-08 05:57:05 -04:00
worker_types_env = environ.get("SYNAPSE_WORKER_TYPES", "").strip()
if not worker_types_env:
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
# No workers, just the main process
worker_types = []
else:
Strip whitespace from worker types in Dockerfile-workers (#14165)
2022-10-14 15:38:04 -04:00
# Split type names by comma, ignoring whitespace.
worker_types = [x.strip() for x in worker_types_env.split(",")]
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
# Create the worker configuration directory if it doesn't already exist
os.makedirs("/conf/workers", exist_ok=True)
# Start worker ports from this arbitrary port
worker_port = 18009
# A counter of worker_type -> int. Used for determining the name for a given
# worker type when generating its config file, as each worker's name is just
# worker_type + instance #
Add missing type definitions for scripts in docker folder (#12280) Signed-off-by: Jorge Florian <jafn28@gmail.com>
2022-04-08 06:10:58 -04:00
worker_type_counter: Dict[str, int] = {}
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
Fix up healthcheck generation for workers docker image (#12405) This wasn't quite generating the right thing.
2022-04-11 09:38:58 -04:00
# A list of internal endpoints to healthcheck, starting with the main process
# which exists even if no workers do.
healthcheck_urls = ["http://localhost:8080/health"]
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
# For each worker type specified by the user, create config values
for worker_type in worker_types:
worker_config = WORKERS_CONFIG.get(worker_type)
if worker_config:
worker_config = worker_config.copy()
else:
log(worker_type + " is an unknown worker type! It will be ignored")
continue
new_worker_count = worker_type_counter.setdefault(worker_type, 0) + 1
worker_type_counter[worker_type] = new_worker_count
# Name workers by their type concatenated with an incrementing number
# e.g. federation_reader1
worker_name = worker_type + str(new_worker_count)
worker_config.update(
Add missing type definitions for scripts in docker folder (#12280) Signed-off-by: Jorge Florian <jafn28@gmail.com>
2022-04-08 06:10:58 -04:00
{"name": worker_name, "port": str(worker_port), "config_path": config_path}
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
)
# Update the shared config with any worker-type specific options
shared_config.update(worker_config["shared_extra_conf"])
Fix up healthcheck generation for workers docker image (#12405) This wasn't quite generating the right thing.
2022-04-11 09:38:58 -04:00
healthcheck_urls.append("http://localhost:%d/health" % (worker_port,))
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
# Check if more than one instance of this worker type has been specified
worker_type_total_count = worker_types.count(worker_type)
if worker_type_total_count > 1:
# Update the shared config with sharding-related options if necessary
add_sharding_to_shared_config(
shared_config, worker_type, worker_name, worker_port
)
# Enable the worker in supervisord
Refactor the Dockerfile-workers configuration script to use Jinja2 templates in Synapse workers' Supervisord blocks. (#13054) Co-authored-by: Richard van der Hoff <1389908+richvdh@users.noreply.github.com>
2022-06-27 06:43:20 -04:00
worker_descriptors.append(worker_config)
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
# Add nginx location blocks for this worker's endpoints (if any are defined)
for pattern in worker_config["endpoint_patterns"]:
# Determine whether we need to load-balance this worker
if worker_type_total_count > 1:
# Create or add to a load-balanced upstream for this worker
nginx_upstreams.setdefault(worker_type, set()).add(worker_port)
# Upstreams are named after the worker_type
upstream = "http://" + worker_type
else:
upstream = "http://localhost:%d" % (worker_port,)
# Note that this endpoint should proxy to this upstream
nginx_locations[pattern] = upstream
# Write out the worker's logging config file
Dockerfile-workers: give the master its own log config (#12466) When we run a worker-mode synapse under docker, everything gets logged to stdout. Currently, output from the workers is tacked with a worker name, for example: ``` 2022-04-13 15:27:56,810 - worker:frontend_proxy1 - synapse.util.caches.lrucache - 154 - INFO - LruCache._expire_old_entries-0 - Dropped 0 items from caches ``` - note `worker:frontend_proxy1`. No such tag is applied to log lines from the master, which makes somewhat confusing reading. To fix this, we generate a dedicated log config file for the master in the same way that we do for the workers, and use that.
2022-04-13 15:50:08 -04:00
log_config_filepath = generate_worker_log_config(environ, worker_name, data_dir)
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
# Then a worker config file
convert(
"/conf/worker.yaml.j2",
"/conf/workers/{name}.yaml".format(name=worker_name),
**worker_config,
worker_log_config_filepath=log_config_filepath,
)
worker_port += 1
# Build the nginx location config blocks
nginx_location_config = ""
for endpoint, upstream in nginx_locations.items():
nginx_location_config += NGINX_LOCATION_CONFIG_BLOCK.format(
endpoint=endpoint,
upstream=upstream,
)
# Determine the load-balancing upstreams to configure
nginx_upstream_config = ""
Create healthcheck script for synapse-workers container (#11429) The intent is to iterate through all the worker ports and only report healthy when all are healthy, starting with the main process.
2021-11-26 09:05:20 -05:00
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
for upstream_worker_type, upstream_worker_ports in nginx_upstreams.items():
body = ""
for port in upstream_worker_ports:
body += " server localhost:%d;\n" % (port,)
# Add to the list of configured upstreams
nginx_upstream_config += NGINX_UPSTREAM_CONFIG_BLOCK.format(
upstream_worker_type=upstream_worker_type,
body=body,
)
# Finally, we'll write out the config files.
Dockerfile-workers: give the master its own log config (#12466) When we run a worker-mode synapse under docker, everything gets logged to stdout. Currently, output from the workers is tacked with a worker name, for example: ``` 2022-04-13 15:27:56,810 - worker:frontend_proxy1 - synapse.util.caches.lrucache - 154 - INFO - LruCache._expire_old_entries-0 - Dropped 0 items from caches ``` - note `worker:frontend_proxy1`. No such tag is applied to log lines from the master, which makes somewhat confusing reading. To fix this, we generate a dedicated log config file for the master in the same way that we do for the workers, and use that.
2022-04-13 15:50:08 -04:00
# log config for the master process
master_log_config = generate_worker_log_config(environ, "master", data_dir)
shared_config["log_config"] = master_log_config
Support registering Application Services when running with workers under Complement. (#12826) Co-authored-by: Patrick Cloke <clokep@users.noreply.github.com>
2022-05-23 09:11:06 -04:00
# Find application service registrations
appservice_registrations = None
appservice_registration_dir = os.environ.get("SYNAPSE_AS_REGISTRATION_DIR")
if appservice_registration_dir:
# Scan for all YAML files that should be application service registrations.
appservice_registrations = [
str(reg_path.resolve())
for reg_path in Path(appservice_registration_dir).iterdir()
if reg_path.suffix.lower() in (".yaml", ".yml")
]
Merge the Complement testing Docker images into a single, multi-purpose image. (#12881) Co-authored-by: Richard van der Hoff <1389908+richvdh@users.noreply.github.com>
2022-06-08 05:57:05 -04:00
workers_in_use = len(worker_types) > 0
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
# Shared homeserver config
convert(
"/conf/shared.yaml.j2",
"/conf/workers/shared.yaml",
shared_worker_config=yaml.dump(shared_config),
Support registering Application Services when running with workers under Complement. (#12826) Co-authored-by: Patrick Cloke <clokep@users.noreply.github.com>
2022-05-23 09:11:06 -04:00
appservice_registrations=appservice_registrations,
Merge the Complement testing Docker images into a single, multi-purpose image. (#12881) Co-authored-by: Richard van der Hoff <1389908+richvdh@users.noreply.github.com>
2022-06-08 05:57:05 -04:00
enable_redis=workers_in_use,
workers_in_use=workers_in_use,
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
)
# Nginx config
convert(
"/conf/nginx.conf.j2",
"/etc/nginx/conf.d/matrix-synapse.conf",
worker_locations=nginx_location_config,
upstream_directives=nginx_upstream_config,
Remove Caddy from the Synapse workers image used in Complement. (#12818)
2022-05-23 05:29:24 -04:00
tls_cert_path=os.environ.get("SYNAPSE_TLS_CERT"),
tls_key_path=os.environ.get("SYNAPSE_TLS_KEY"),
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
)
# Supervisord config
Dockerfile-workers: reduce the amount we install (#12464) This is an attempt to reduce the rebuild time. In short, we reduce the amount of stuff that the dockerfile installs, so as to give a faster startup.
2022-04-14 10:36:49 -04:00
os.makedirs("/etc/supervisor", exist_ok=True)
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
convert(
"/conf/supervisord.conf.j2",
Dockerfile-workers: reduce the amount we install (#12464) This is an attempt to reduce the rebuild time. In short, we reduce the amount of stuff that the dockerfile installs, so as to give a faster startup.
2022-04-14 10:36:49 -04:00
"/etc/supervisor/supervisord.conf",
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
main_config_path=config_path,
Merge the Complement testing Docker images into a single, multi-purpose image. (#12881) Co-authored-by: Richard van der Hoff <1389908+richvdh@users.noreply.github.com>
2022-06-08 05:57:05 -04:00
enable_redis=workers_in_use,
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
)
Refactor the Dockerfile-workers configuration script to use Jinja2 templates in Synapse workers' Supervisord blocks. (#13054) Co-authored-by: Richard van der Hoff <1389908+richvdh@users.noreply.github.com>
2022-06-27 06:43:20 -04:00
convert(
"/conf/synapse.supervisord.conf.j2",
"/etc/supervisor/conf.d/synapse.conf",
workers=worker_descriptors,
main_config_path=config_path,
Improve startup times in Complement test runs against workers, particularly in CPU-constrained environments. (#13127) Co-authored-by: Richard van der Hoff <1389908+richvdh@users.noreply.github.com>
2022-06-30 07:58:12 -04:00
use_forking_launcher=environ.get("SYNAPSE_USE_EXPERIMENTAL_FORKING_LAUNCHER"),
Refactor the Dockerfile-workers configuration script to use Jinja2 templates in Synapse workers' Supervisord blocks. (#13054) Co-authored-by: Richard van der Hoff <1389908+richvdh@users.noreply.github.com>
2022-06-27 06:43:20 -04:00
)
Create healthcheck script for synapse-workers container (#11429) The intent is to iterate through all the worker ports and only report healthy when all are healthy, starting with the main process.
2021-11-26 09:05:20 -05:00
# healthcheck config
convert(
"/conf/healthcheck.sh.j2",
"/healthcheck.sh",
healthcheck_urls=healthcheck_urls,
)
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
# Ensure the logging directory exists
log_dir = data_dir + "/logs"
if not os.path.exists(log_dir):
os.mkdir(log_dir)
Dockerfile-workers: give the master its own log config (#12466) When we run a worker-mode synapse under docker, everything gets logged to stdout. Currently, output from the workers is tacked with a worker name, for example: ``` 2022-04-13 15:27:56,810 - worker:frontend_proxy1 - synapse.util.caches.lrucache - 154 - INFO - LruCache._expire_old_entries-0 - Dropped 0 items from caches ``` - note `worker:frontend_proxy1`. No such tag is applied to log lines from the master, which makes somewhat confusing reading. To fix this, we generate a dedicated log config file for the master in the same way that we do for the workers, and use that.
2022-04-13 15:50:08 -04:00
def generate_worker_log_config(
environ: Mapping[str, str], worker_name: str, data_dir: str
) -> str:
"""Generate a log.config file for the given worker.
Returns: the path to the generated file
"""
# Check whether we should write worker logs to disk, in addition to the console
Add the ability to set the log level using the `SYNAPSE_TEST_LOG_LEVEL` environment when using `complement.sh`. (#13152)
2022-07-05 05:46:20 -04:00
extra_log_template_args: Dict[str, Optional[str]] = {}
Dockerfile-workers: give the master its own log config (#12466) When we run a worker-mode synapse under docker, everything gets logged to stdout. Currently, output from the workers is tacked with a worker name, for example: ``` 2022-04-13 15:27:56,810 - worker:frontend_proxy1 - synapse.util.caches.lrucache - 154 - INFO - LruCache._expire_old_entries-0 - Dropped 0 items from caches ``` - note `worker:frontend_proxy1`. No such tag is applied to log lines from the master, which makes somewhat confusing reading. To fix this, we generate a dedicated log config file for the master in the same way that we do for the workers, and use that.
2022-04-13 15:50:08 -04:00
if environ.get("SYNAPSE_WORKERS_WRITE_LOGS_TO_DISK"):
Add the ability to set the log level using the `SYNAPSE_TEST_LOG_LEVEL` environment when using `complement.sh`. (#13152)
2022-07-05 05:46:20 -04:00
extra_log_template_args["LOG_FILE_PATH"] = f"{data_dir}/logs/{worker_name}.log"
extra_log_template_args["SYNAPSE_LOG_LEVEL"] = environ.get("SYNAPSE_LOG_LEVEL")
extra_log_template_args["SYNAPSE_LOG_SENSITIVE"] = environ.get(
"SYNAPSE_LOG_SENSITIVE"
)
Dockerfile-workers: give the master its own log config (#12466) When we run a worker-mode synapse under docker, everything gets logged to stdout. Currently, output from the workers is tacked with a worker name, for example: ``` 2022-04-13 15:27:56,810 - worker:frontend_proxy1 - synapse.util.caches.lrucache - 154 - INFO - LruCache._expire_old_entries-0 - Dropped 0 items from caches ``` - note `worker:frontend_proxy1`. No such tag is applied to log lines from the master, which makes somewhat confusing reading. To fix this, we generate a dedicated log config file for the master in the same way that we do for the workers, and use that.
2022-04-13 15:50:08 -04:00
# Render and write the file
Add the ability to set the log level using the `SYNAPSE_TEST_LOG_LEVEL` environment when using `complement.sh`. (#13152)
2022-07-05 05:46:20 -04:00
log_config_filepath = f"/conf/workers/{worker_name}.log.config"
Dockerfile-workers: give the master its own log config (#12466) When we run a worker-mode synapse under docker, everything gets logged to stdout. Currently, output from the workers is tacked with a worker name, for example: ``` 2022-04-13 15:27:56,810 - worker:frontend_proxy1 - synapse.util.caches.lrucache - 154 - INFO - LruCache._expire_old_entries-0 - Dropped 0 items from caches ``` - note `worker:frontend_proxy1`. No such tag is applied to log lines from the master, which makes somewhat confusing reading. To fix this, we generate a dedicated log config file for the master in the same way that we do for the workers, and use that.
2022-04-13 15:50:08 -04:00
convert(
"/conf/log.config",
log_config_filepath,
worker_name=worker_name,
**extra_log_template_args,
Improve startup times in Complement test runs against workers, particularly in CPU-constrained environments. (#13127) Co-authored-by: Richard van der Hoff <1389908+richvdh@users.noreply.github.com>
2022-06-30 07:58:12 -04:00
include_worker_name_in_log_line=environ.get(
"SYNAPSE_USE_EXPERIMENTAL_FORKING_LAUNCHER"
),
Dockerfile-workers: give the master its own log config (#12466) When we run a worker-mode synapse under docker, everything gets logged to stdout. Currently, output from the workers is tacked with a worker name, for example: ``` 2022-04-13 15:27:56,810 - worker:frontend_proxy1 - synapse.util.caches.lrucache - 154 - INFO - LruCache._expire_old_entries-0 - Dropped 0 items from caches ``` - note `worker:frontend_proxy1`. No such tag is applied to log lines from the master, which makes somewhat confusing reading. To fix this, we generate a dedicated log config file for the master in the same way that we do for the workers, and use that.
2022-04-13 15:50:08 -04:00
)
return log_config_filepath
disallow-untyped-defs in `docker` and `stubs` directories (#12528)
2022-04-25 08:32:35 -04:00
def main(args: List[str], environ: MutableMapping[str, str]) -> None:
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
config_dir = environ.get("SYNAPSE_CONFIG_DIR", "/data")
config_path = environ.get("SYNAPSE_CONFIG_PATH", config_dir + "/homeserver.yaml")
data_dir = environ.get("SYNAPSE_DATA_DIR", "/data")
# override SYNAPSE_NO_TLS, we don't support TLS in worker mode,
# this needs to be handled by a frontend proxy
environ["SYNAPSE_NO_TLS"] = "yes"
# Generate the base homeserver config if one does not yet exist
if not os.path.exists(config_path):
log("Generating base homeserver config")
generate_base_homeserver_config()
# This script may be run multiple times (mostly by Complement, see note at top of file).
# Don't re-configure workers in this instance.
mark_filepath = "/conf/workers_have_been_configured"
if not os.path.exists(mark_filepath):
# Always regenerate all other config files
generate_worker_files(environ, config_path, data_dir)
# Mark workers as being configured
with open(mark_filepath, "w") as f:
f.write("")
Set LD_PRELOAD to load jemalloc in Dockerfile-workers. (#14182)
2022-10-14 08:29:49 -04:00
# Lifted right out of start.py
jemallocpath = "/usr/lib/%s-linux-gnu/libjemalloc.so.2" % (platform.machine(),)
if os.path.isfile(jemallocpath):
environ["LD_PRELOAD"] = jemallocpath
else:
log("Could not find %s, will not use" % (jemallocpath,))
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
# Start supervisord, which will start Synapse, all of the configured worker
# processes, redis, nginx etc. according to the config we created above.
Dockerfile-workers: reduce the amount we install (#12464) This is an attempt to reduce the rebuild time. In short, we reduce the amount of stuff that the dockerfile installs, so as to give a faster startup.
2022-04-14 10:36:49 -04:00
log("Starting supervisord")
Set LD_PRELOAD to load jemalloc in Dockerfile-workers. (#14182)
2022-10-14 08:29:49 -04:00
os.execle(
Dockerfile-workers: reduce the amount we install (#12464) This is an attempt to reduce the rebuild time. In short, we reduce the amount of stuff that the dockerfile installs, so as to give a faster startup.
2022-04-14 10:36:49 -04:00
"/usr/local/bin/supervisord",
"supervisord",
"-c",
"/etc/supervisor/supervisord.conf",
Set LD_PRELOAD to load jemalloc in Dockerfile-workers. (#14182)
2022-10-14 08:29:49 -04:00
environ,
Dockerfile-workers: reduce the amount we install (#12464) This is an attempt to reduce the rebuild time. In short, we reduce the amount of stuff that the dockerfile installs, so as to give a faster startup.
2022-04-14 10:36:49 -04:00
)
Add a dockerfile for running a set of Synapse worker processes (#9162) This PR adds a Dockerfile and some supporting files to the `docker/` directory. The Dockerfile's intention is to spin up a container with: * A Synapse main process. * Any desired worker processes, defined by a `SYNAPSE_WORKERS` environment variable supplied at runtime. * A redis for worker communication. * A nginx for routing traffic. * A supervisord to start all worker processes and monitor them if any go down. Note that **this is not currently intended to be used in production**. If you'd like to use Synapse workers with Docker, instead make use of the official image, with one worker per container. The purpose of this dockerfile is currently to allow testing Synapse in worker mode with the [Complement](https://github.com/matrix-org/complement/) test suite. `configure_workers_and_start.py` is where most of the magic happens in this PR. It reads from environment variables (documented in the file) and creates all necessary config files for the processes. It is the entrypoint of the Dockerfile, and thus is run any time the docker container is spun up, recreating all config files in case you want to use a different set of workers. One can specify which workers they'd like to use by setting the `SYNAPSE_WORKERS` environment variable (as a comma-separated list of arbitrary worker names) or by setting it to `*` for all worker processes. We will be using the latter in CI. Huge thanks to @MatMaul for helping get this all working :tada: This PR is paired with its equivalent on the Complement side: https://github.com/matrix-org/complement/pull/62. Note, for the purpose of testing this PR before it's merged: You'll need to (re)build the base Synapse docker image for everything to work (`matrixdotorg/synapse:latest`). Then build the worker-based docker image on top (`matrixdotorg/synapse:workers`).
2021-04-14 08:54:49 -04:00
if __name__ == "__main__":
main(sys.argv, os.environ)
Reference in New Issue Copy Permalink