Merge pull request #2812 from matrix-org/erikj/media_storage_provider_config

Make storage providers configurable

synapse/config/repository.py

@@ -16,6 +16,8 @@
 from ._base import Config, ConfigError
 from collections import namedtuple
 
+from synapse.util.module_loader import load_module
+
 
 MISSING_NETADDR = (
     "Missing netaddr library. This is required for URL preview API."
@@ -36,6 +38,14 @@ ThumbnailRequirement = namedtuple(
     "ThumbnailRequirement", ["width", "height", "method", "media_type"]
 )
 
+MediaStorageProviderConfig = namedtuple(
+    "MediaStorageProviderConfig", (
+        "store_local",  # Whether to store newly uploaded local files
+        "store_remote",  # Whether to store newly downloaded remote files
+        "store_synchronous",  # Whether to wait for successful storage for local uploads
+    ),
+)
+
 
 def parse_thumbnail_requirements(thumbnail_sizes):
     """ Takes a list of dictionaries with "width", "height", and "method" keys
@@ -73,14 +83,59 @@ class ContentRepositoryConfig(Config):
 
         self.media_store_path = self.ensure_directory(config["media_store_path"])
 
-        self.backup_media_store_path = config.get("backup_media_store_path")
+        backup_media_store_path = config.get("backup_media_store_path")
-        if self.backup_media_store_path:
+
-            self.backup_media_store_path = self.ensure_directory(
+        synchronous_backup_media_store = config.get(
-                self.backup_media_store_path
+            "synchronous_backup_media_store", False
         )
 
-        self.synchronous_backup_media_store = config.get(
+        storage_providers = config.get("media_storage_providers", [])
-            "synchronous_backup_media_store", False
+
+        if backup_media_store_path:
+            if storage_providers:
+                raise ConfigError(
+                    "Cannot use both 'backup_media_store_path' and 'storage_providers'"
+                )
+
+            storage_providers = [{
+                "module": "file_system",
+                "store_local": True,
+                "store_synchronous": synchronous_backup_media_store,
+                "store_remote": True,
+                "config": {
+                    "directory": backup_media_store_path,
+                }
+            }]
+
+        # This is a list of config that can be used to create the storage
+        # providers. The entries are tuples of (Class, class_config,
+        # MediaStorageProviderConfig), where Class is the class of the provider,
+        # the class_config the config to pass to it, and
+        # MediaStorageProviderConfig are options for StorageProviderWrapper.
+        #
+        # We don't create the storage providers here as not all workers need
+        # them to be started.
+        self.media_storage_providers = []
+
+        for provider_config in storage_providers:
+            # We special case the module "file_system" so as not to need to
+            # expose FileStorageProviderBackend
+            if provider_config["module"] == "file_system":
+                provider_config["module"] = (
+                    "synapse.rest.media.v1.storage_provider"
+                    ".FileStorageProviderBackend"
+                )
+
+            provider_class, parsed_config = load_module(provider_config)
+
+            wrapper_config = MediaStorageProviderConfig(
+                provider_config.get("store_local", False),
+                provider_config.get("store_remote", False),
+                provider_config.get("store_synchronous", False),
+            )
+
+            self.media_storage_providers.append(
+                (provider_class, parsed_config, wrapper_config,)
             )
 
         self.uploads_path = self.ensure_directory(config["uploads_path"])
@@ -127,13 +182,19 @@ class ContentRepositoryConfig(Config):
         # Directory where uploaded images and attachments are stored.
         media_store_path: "%(media_store)s"
 
-        # A secondary directory where uploaded images and attachments are
+        # Media storage providers allow media to be stored in different
-        # stored as a backup.
+        # locations.
-        # backup_media_store_path: "%(media_store)s"
+        # media_storage_providers:
-
+        # - module: file_system
-        # Whether to wait for successful write to backup media store before
+        #   # Whether to write new local files.
-        # returning successfully.
+        #   store_local: false
-        # synchronous_backup_media_store: false
+        #   # Whether to write new remote media
+        #   store_remote: false
+        #   # Whether to block upload requests waiting for write to this
+        #   # provider to complete
+        #   store_synchronous: false
+        #   config:
+        #     directory: /mnt/some/other/directory
 
         # Directory where in-progress uploads are stored.
         uploads_path: "%(uploads_path)s"

synapse/rest/media/v1/media_repository.py

@@ -27,9 +27,7 @@ from .identicon_resource import IdenticonResource
 from .preview_url_resource import PreviewUrlResource
 from .filepath import MediaFilePaths
 from .thumbnailer import Thumbnailer
-from .storage_provider import (
+from .storage_provider import StorageProviderWrapper
-    StorageProviderWrapper, FileStorageProviderBackend,
-)
 from .media_storage import MediaStorage
 
 from synapse.http.matrixfederationclient import MatrixFederationHttpClient
@@ -81,17 +79,13 @@ class MediaRepository(object):
         # potentially upload to.
         storage_providers = []
 
-        # TODO: Move this into config and allow other storage providers to be
+        for clz, provider_config, wrapper_config in hs.config.media_storage_providers:
-        # defined.
+            backend = clz(hs, provider_config)
-        if hs.config.backup_media_store_path:
-            backend = FileStorageProviderBackend(
-                self.primary_base_path, hs.config.backup_media_store_path,
-            )
             provider = StorageProviderWrapper(
                 backend,
-                store=True,
+                store_local=wrapper_config.store_local,
-                store_synchronous=hs.config.synchronous_backup_media_store,
+                store_remote=wrapper_config.store_remote,
-                store_remote=True,
+                store_synchronous=wrapper_config.store_synchronous,
             )
             storage_providers.append(provider)
 

synapse/rest/media/v1/storage_provider.py

@@ -17,6 +17,7 @@ from twisted.internet import defer, threads
 
 from .media_storage import FileResponder
 
+from synapse.config._base import Config
 from synapse.util.logcontext import preserve_fn
 
 import logging
@@ -64,14 +65,14 @@ class StorageProviderWrapper(StorageProvider):
 
     Args:
         backend (StorageProvider)
-        store (bool): Whether to store new files or not.
+        store_local (bool): Whether to store new local files or not.
         store_synchronous (bool): Whether to wait for file to be successfully
             uploaded, or todo the upload in the backgroud.
         store_remote (bool): Whether remote media should be uploaded
     """
-    def __init__(self, backend, store, store_synchronous, store_remote):
+    def __init__(self, backend, store_local, store_synchronous, store_remote):
         self.backend = backend
-        self.store = store
+        self.store_local = store_local
         self.store_synchronous = store_synchronous
         self.store_remote = store_remote
 
@@ -97,13 +98,13 @@ class FileStorageProviderBackend(StorageProvider):
     """A storage provider that stores files in a directory on a filesystem.
 
     Args:
-        cache_directory (str): Base path of the local media repository
+        hs (HomeServer)
-        base_directory (str): Base path to store new files
+        config: The config returned by `parse_config`.
     """
 
-    def __init__(self, cache_directory, base_directory):
+    def __init__(self, hs, config):
-        self.cache_directory = cache_directory
+        self.cache_directory = hs.config.media_store_path
-        self.base_directory = base_directory
+        self.base_directory = config
 
     def store_file(self, path, file_info):
         """See StorageProvider.store_file"""
@@ -125,3 +126,15 @@ class FileStorageProviderBackend(StorageProvider):
         backup_fname = os.path.join(self.base_directory, path)
         if os.path.isfile(backup_fname):
             return FileResponder(open(backup_fname, "rb"))
+
+    @staticmethod
+    def parse_config(config):
+        """Called on startup to parse config supplied. This should parse
+        the config and raise if there is a problem.
+
+        The returned value is passed into the constructor.
+
+        In this case we only care about a single param, the directory, so let's
+        just pull that out.
+        """
+        return Config.ensure_directory(config["directory"])