Merge pull request #2701 from matrix-org/rav/one_mediarepo_to_rule_them_all

Try to avoid having multiple PreviewUrlResource instances

6 files changed, 48 insertions, 13 deletions

diff --git a/docs/workers.rst b/docs/workers.rst
index 3cc8b3d82e..b39f79058e 100644
--- a/docs/workers.rst
+++ b/docs/workers.rst
@@ -170,6 +170,10 @@ Handles the media repository. It can handle all endpoints starting with::
 
     /_matrix/media/
 
+You should also set ``enable_media_repo: False`` in the shared configuration
+file to stop the main synapse running background jobs related to managing the
+media repository.
+
 Note this worker cannot be load-balanced: only one instance should be active.
 
 ``synapse.app.client_reader``
@@ -203,4 +207,3 @@ the ``worker_main_http_uri`` setting in the frontend_proxy worker configuration
 file. For example::
 
     worker_main_http_uri: http://127.0.0.1:8008
-
diff --git a/synapse/app/homeserver.py b/synapse/app/homeserver.py
index 9e26146338..6b8875afb4 100755
--- a/synapse/app/homeserver.py
+++ b/synapse/app/homeserver.py
@@ -43,7 +43,6 @@ from synapse.rest import ClientRestResource
 from synapse.rest.key.v1.server_key_resource import LocalKey
 from synapse.rest.key.v2 import KeyApiV2Resource
 from synapse.rest.media.v0.content_repository import ContentRepoResource
-from synapse.rest.media.v1.media_repository import MediaRepositoryResource
 from synapse.server import HomeServer
 from synapse.storage import are_all_users_on_domain
 from synapse.storage.engines import IncorrectDatabaseSetup, create_engine
@@ -195,14 +194,19 @@ class SynapseHomeServer(HomeServer):
             })
 
         if name in ["media", "federation", "client"]:
-            media_repo = MediaRepositoryResource(self)
-            resources.update({
-                MEDIA_PREFIX: media_repo,
-                LEGACY_MEDIA_PREFIX: media_repo,
-                CONTENT_REPO_PREFIX: ContentRepoResource(
-                    self, self.config.uploads_path
-                ),
-            })
+            if self.get_config().enable_media_repo:
+                media_repo = self.get_media_repository_resource()
+                resources.update({
+                    MEDIA_PREFIX: media_repo,
+                    LEGACY_MEDIA_PREFIX: media_repo,
+                    CONTENT_REPO_PREFIX: ContentRepoResource(
+                        self, self.config.uploads_path
+                    ),
+                })
+            elif name == "media":
+                raise ConfigError(
+                    "'media' resource conflicts with enable_media_repo=False",
+                )
 
         if name in ["keys", "federation"]:
             resources.update({
diff --git a/synapse/app/media_repository.py b/synapse/app/media_repository.py
index 36c18bdbcb..c4e5f0965d 100644
--- a/synapse/app/media_repository.py
+++ b/synapse/app/media_repository.py
@@ -35,7 +35,6 @@ from synapse.replication.slave.storage.registration import SlavedRegistrationSto
 from synapse.replication.slave.storage.transactions import TransactionStore
 from synapse.replication.tcp.client import ReplicationClientHandler
 from synapse.rest.media.v0.content_repository import ContentRepoResource
-from synapse.rest.media.v1.media_repository import MediaRepositoryResource
 from synapse.server import HomeServer
 from synapse.storage.engines import create_engine
 from synapse.storage.media_repository import MediaRepositoryStore
@@ -89,7 +88,7 @@ class MediaRepositoryServer(HomeServer):
                 if name == "metrics":
                     resources[METRICS_PREFIX] = MetricsResource(self)
                 elif name == "media":
-                    media_repo = MediaRepositoryResource(self)
+                    media_repo = self.get_media_repository_resource()
                     resources.update({
                         MEDIA_PREFIX: media_repo,
                         LEGACY_MEDIA_PREFIX: media_repo,
@@ -151,6 +150,13 @@ def start(config_options):
 
     assert config.worker_app == "synapse.app.media_repository"
 
+    if config.enable_media_repo:
+        _base.quit_with_error(
+            "enable_media_repo must be disabled in the main synapse process\n"
+            "before the media repo can be run in a separate worker.\n"
+            "Please add ``enable_media_repo: false`` to the main config\n"
+        )
+
     setup_logging(config, use_worker_options=True)
 
     events.USE_FROZEN_DICTS = config.use_frozen_dicts
diff --git a/synapse/config/server.py b/synapse/config/server.py
index 4d9193536d..edb90a1348 100644
--- a/synapse/config/server.py
+++ b/synapse/config/server.py
@@ -41,6 +41,12 @@ class ServerConfig(Config):
         # false only if we are updating the user directory in a worker
         self.update_user_directory = config.get("update_user_directory", True)
 
+        # whether to enable the media repository endpoints. This should be set
+        # to false if the media repository is running as a separate endpoint;
+        # doing so ensures that we will not run cache cleanup jobs on the
+        # master, potentially causing inconsistency.
+        self.enable_media_repo = config.get("enable_media_repo", True)
+
         self.filter_timeline_limit = config.get("filter_timeline_limit", -1)
 
         # Whether we should block invites sent to users on this server
diff --git a/synapse/server.py b/synapse/server.py
index 4746cc7b6c..853f4647b7 100644
--- a/synapse/server.py
+++ b/synapse/server.py
@@ -60,7 +60,10 @@ from synapse.http.matrixfederationclient import MatrixFederationHttpClient
 from synapse.notifier import Notifier
 from synapse.push.action_generator import ActionGenerator
 from synapse.push.pusherpool import PusherPool
-from synapse.rest.media.v1.media_repository import MediaRepository
+from synapse.rest.media.v1.media_repository import (
+    MediaRepository,
+    MediaRepositoryResource,
+)
 from synapse.state import StateHandler
 from synapse.storage import DataStore
 from synapse.streams.events import EventSources
@@ -121,6 +124,7 @@ class HomeServer(object):
         'http_client_context_factory',
         'simple_http_client',
         'media_repository',
+        'media_repository_resource',
         'federation_transport_client',
         'federation_sender',
         'receipts_handler',
@@ -293,6 +297,11 @@ class HomeServer(object):
             **self.db_config.get("args", {})
         )
 
+    def build_media_repository_resource(self):
+        # build the media repo resource. This indirects through the HomeServer
+        # to ensure that we only have a single instance of
+        return MediaRepositoryResource(self)
+
     def build_media_repository(self):
         return MediaRepository(self)
 
diff --git a/synapse/server.pyi b/synapse/server.pyi
index e8c0386b7f..3064a497eb 100644
--- a/synapse/server.pyi
+++ b/synapse/server.pyi
@@ -5,6 +5,7 @@ import synapse.handlers
 import synapse.handlers.auth
 import synapse.handlers.device
 import synapse.handlers.e2e_keys
+import synapse.rest.media.v1.media_repository
 import synapse.storage
 import synapse.state
 
@@ -35,3 +36,9 @@ class HomeServer(object):
 
     def get_federation_transport_client(self) -> synapse.federation.transport.client.TransportLayerClient:
         pass
+
+    def get_media_repository_resource(self) -> synapse.rest.media.v1.media_repository.MediaRepositoryResource:
+        pass
+
+    def get_media_repository(self) -> synapse.rest.media.v1.media_repository.MediaRepository:
+        pass