Docker: avoid changing userid unnecessarily (#11209)

* Docker image: avoid changing user during `generate`

The intention was always that the config files get written as the initial user
(normally root) - only the data directory needs to be writable by Synapse. This
got changed in https://github.com/matrix-org/synapse/pull/5970, but that seems
to have been a mistake.

* Avoid changing user if no explicit UID is given

* changelog

changelog.d/11209.docker

@@ -0,0 +1 @@
+Avoid changing userid when started as a non-root user, and no explicit `UID` is set.

docker/README.md

@@ -65,7 +65,8 @@ The following environment variables are supported in `generate` mode:
 * `SYNAPSE_DATA_DIR`: where the generated config will put persistent data
   such as the database and media store. Defaults to `/data`.
 * `UID`, `GID`: the user id and group id to use for creating the data
-  directories. Defaults to `991`, `991`.
+  directories. If unset, and no user is set via `docker run --user`, defaults
+  to `991`, `991`.
 
 ## Running synapse
 
@@ -97,7 +98,9 @@ The following environment variables are supported in `run` mode:
   `<SYNAPSE_CONFIG_DIR>/homeserver.yaml`.
 * `SYNAPSE_WORKER`: module to execute, used when running synapse with workers.
    Defaults to `synapse.app.homeserver`, which is suitable for non-worker mode.
-* `UID`, `GID`: the user and group id to run Synapse as. Defaults to `991`, `991`.
+* `UID`, `GID`: the user and group id to run Synapse as. If unset, and no user
+  is set via `docker run --user`, defaults to `991`, `991`. Note that this user
+  must have permission to read the config files, and write to the data directories.
 * `TZ`: the [timezone](https://en.wikipedia.org/wiki/List_of_tz_database_time_zones) the container will run with. Defaults to `UTC`.
 
 For more complex setups (e.g. for workers) you can also pass your args directly to synapse using `run` mode. For example like this:
@@ -186,7 +189,7 @@ point to another Dockerfile.
 ## Disabling the healthcheck
 
 If you are using a non-standard port or tls inside docker you can disable the healthcheck
-whilst running the above `docker run` commands. 
+whilst running the above `docker run` commands.
 
 ```
    --no-healthcheck
@@ -212,7 +215,7 @@ If you wish to point the healthcheck at a different port with docker command, ad
 ## Setting the healthcheck in docker-compose file
 
 You can add the following to set a custom healthcheck in a docker compose file.
-You will need docker-compose version >2.1 for this to work. 
+You will need docker-compose version >2.1 for this to work.
 
 ```
 healthcheck:
@@ -226,5 +229,5 @@ healthcheck:
 ## Using jemalloc
 
 Jemalloc is embedded in the image and will be used instead of the default allocator.
-You can read about jemalloc by reading the Synapse 
+You can read about jemalloc by reading the Synapse
 [README](https://github.com/matrix-org/synapse/blob/HEAD/README.rst#help-synapse-is-slow-and-eats-all-my-ram-cpu).

docker/start.py

@@ -120,6 +120,7 @@ def generate_config_from_template(config_dir, config_path, environ, ownership):
     ]
 
     if ownership is not None:
+        log(f"Setting ownership on /data to {ownership}")
         subprocess.check_output(["chown", "-R", ownership, "/data"])
         args = ["gosu", ownership] + args
 
@@ -144,12 +145,18 @@ def run_generate_config(environ, ownership):
     config_path = environ.get("SYNAPSE_CONFIG_PATH", config_dir + "/homeserver.yaml")
     data_dir = environ.get("SYNAPSE_DATA_DIR", "/data")
 
+    if ownership is not None:
+        # make sure that synapse has perms to write to the data dir.
+        log(f"Setting ownership on {data_dir} to {ownership}")
+        subprocess.check_output(["chown", ownership, data_dir])
+
     # create a suitable log config from our template
     log_config_file = "%s/%s.log.config" % (config_dir, server_name)
     if not os.path.exists(log_config_file):
         log("Creating log config %s" % (log_config_file,))
         convert("/conf/log.config", log_config_file, environ)
 
+    # generate the main config file, and a signing key.
     args = [
         "python",
         "-m",
@@ -168,29 +175,23 @@ def run_generate_config(environ, ownership):
         "--open-private-ports",
     ]
     # log("running %s" % (args, ))
-
+    os.execv("/usr/local/bin/python", args)
-    if ownership is not None:
-        # make sure that synapse has perms to write to the data dir.
-        subprocess.check_output(["chown", ownership, data_dir])
-
-        args = ["gosu", ownership] + args
-        os.execv("/usr/sbin/gosu", args)
-    else:
-        os.execv("/usr/local/bin/python", args)
 
 
 def main(args, environ):
     mode = args[1] if len(args) > 1 else "run"
-    desired_uid = int(environ.get("UID", "991"))
-    desired_gid = int(environ.get("GID", "991"))
-    synapse_worker = environ.get("SYNAPSE_WORKER", "synapse.app.homeserver")
-    if (desired_uid == os.getuid()) and (desired_gid == os.getgid()):
-        ownership = None
-    else:
-        ownership = "{}:{}".format(desired_uid, desired_gid)
 
-    if ownership is None:
+    # if we were given an explicit user to switch to, do so
-        log("Will not perform chmod/gosu as UserID already matches request")
+    ownership = None
+    if "UID" in environ:
+        desired_uid = int(environ["UID"])
+        desired_gid = int(environ.get("GID", "991"))
+        ownership = f"{desired_uid}:{desired_gid}"
+    elif os.getuid() == 0:
+        # otherwise, if we are running as root, use user 991
+        ownership = "991:991"
+
+    synapse_worker = environ.get("SYNAPSE_WORKER", "synapse.app.homeserver")
 
     # In generate mode, generate a configuration and missing keys, then exit
     if mode == "generate":