kegsay 11a8ae0632 complement: enable dirty runs (#16520) 5 months ago
..
complement 11a8ae0632 complement: enable dirty runs (#16520) 5 months ago
conf 11a8ae0632 complement: enable dirty runs (#16520) 5 months ago
conf-workers 224ef0b669 Unix Sockets for HTTP Replication (#15708) 8 months ago
Dockerfile 3cf1a3aa17 Use bookwork as docker base image (#16324) 6 months ago
Dockerfile-dhvirtualenv 561d06b481 Remove support for Python 3.7 (#15851) 8 months ago
Dockerfile-workers 3cf1a3aa17 Use bookwork as docker base image (#16324) 6 months ago
README-testing.md 10e4093839 Call out buildkit is required when building test docker images (#13338) 1 year ago
README.md 30a5076da8 Log when events are (unexpectedly) filtered out of responses in tests (#14213) 10 months ago
build_debian.sh d7141e0b8b Fix Shellcheck SC2006: Use $(...) notation 2 years ago
configure_workers_and_start.py 8940d1b28e Add `/notifications` endpoint to workers (#16265) 6 months ago
editable.Dockerfile 3cf1a3aa17 Use bookwork as docker base image (#16324) 6 months ago
prefix-log d743b25c8f Use supervisord to supervise Postgres and Caddy in the Complement image. (#12480) 1 year ago
start.py aa483cb4c9 Update ruff config (#16283) 6 months ago

README-testing.md

Running tests against a dockerised Synapse

It's possible to run integration tests against Synapse using Complement. Complement is a Matrix Spec compliance test suite for homeservers, and supports any homeserver docker image configured to listen on ports 8008/8448. This document contains instructions for building Synapse docker images that can be run inside Complement for testing purposes.

Note that running Synapse's unit tests from within the docker image is not supported.

Using the Complement launch script

scripts-dev/complement.sh is a script that will automatically build and run Synapse against Complement. Consult the contributing guide for instructions on how to use it.

Building and running the images manually

Under some circumstances, you may wish to build the images manually. The instructions below will lead you to doing that.

Note that these images can only be built using BuildKit, therefore BuildKit needs to be enabled when calling docker build. This can be done by setting DOCKER_BUILDKIT=1 in your environment.

Start by building the base Synapse docker image. If you wish to run tests with the latest release of Synapse, instead of your current checkout, you can skip this step. From the root of the repository:

docker build -t matrixdotorg/synapse -f docker/Dockerfile .

Next, build the workerised Synapse docker image, which is a layer over the base image.

docker build -t matrixdotorg/synapse-workers -f docker/Dockerfile-workers .

Finally, build the multi-purpose image for Complement, which is a layer over the workers image.

docker build -t complement-synapse -f docker/complement/Dockerfile docker/complement

This will build an image with the tag complement-synapse, which can be handed to Complement for testing via the COMPLEMENT_BASE_IMAGE environment variable. Refer to Complement's documentation for how to run the tests, as well as the various available command line flags.

See the Complement image README for information about the expected environment variables.

Running the Dockerfile-worker image standalone

For manual testing of a multi-process Synapse instance in Docker, Dockerfile-workers is a Dockerfile that will produce an image bundling all necessary components together for a workerised homeserver instance.

This includes any desired Synapse worker processes, a nginx to route traffic accordingly, a redis for worker communication and a supervisord instance to start up and monitor all processes. You will need to provide your own postgres container to connect to, and TLS is not handled by the container.

Once you've built the image using the above instructions, you can run it. Be sure you've set up a volume according to the usual Synapse docker instructions. Then run something along the lines of:

docker run -d --name synapse \
    --mount type=volume,src=synapse-data,dst=/data \
    -p 8008:8008 \
    -e SYNAPSE_SERVER_NAME=my.matrix.host \
    -e SYNAPSE_REPORT_STATS=no \
    -e POSTGRES_HOST=postgres \
    -e POSTGRES_USER=postgres \
    -e POSTGRES_PASSWORD=somesecret \
    -e SYNAPSE_WORKER_TYPES=synchrotron,media_repository,user_dir \
    -e SYNAPSE_WORKERS_WRITE_LOGS_TO_DISK=1 \
    matrixdotorg/synapse-workers

...substituting POSTGRES* variables for those that match a postgres host you have available (usually a running postgres docker container).

Workers

The SYNAPSE_WORKER_TYPES environment variable is a comma-separated list of workers to use when running the container. All possible worker names are defined by the keys of the WORKERS_CONFIG variable in this script, which the Dockerfile makes use of to generate appropriate worker, nginx and supervisord config files.

Sharding is supported for a subset of workers, in line with the worker documentation. To run multiple instances of a given worker type, simply specify the type multiple times in SYNAPSE_WORKER_TYPES (e.g SYNAPSE_WORKER_TYPES=event_creator,event_creator...).

Otherwise, SYNAPSE_WORKER_TYPES can either be left empty or unset to spawn no workers (leaving only the main process). The container will only be configured to use Redis-based worker mode if there are workers enabled.

Logging

Logs for workers and the main process are logged to stdout and can be viewed with standard docker logs tooling. Worker logs contain their worker name after the timestamp.

Setting SYNAPSE_WORKERS_WRITE_LOGS_TO_DISK=1 will cause worker logs to be written to <data_dir>/logs/<worker_name>.log. Logs are kept for 1 week and rotate every day at 00: 00, according to the container's clock. Logging for the main process must still be configured by modifying the homeserver's log config in your Synapse data volume.

Application Services

Setting the SYNAPSE_AS_REGISTRATION_DIR environment variable to the path of a directory (within the container) will cause the configuration script to scan that directory for .yaml/.yml registration files. Synapse will be configured to load these configuration files.

TLS Termination

Nginx is present in the image to route requests to the appropriate workers, but it does not serve TLS by default.

You can configure SYNAPSE_TLS_CERT and SYNAPSE_TLS_KEY to point to a TLS certificate and key (respectively), both in PEM (textual) format. In this case, Nginx will additionally serve using HTTPS on port 8448.