Add prometheus metrics to track federation delays (#8430)

Add a pair of federation metrics to track the delays in sending PDUs to/from 
particular servers.

4 files changed, 30 insertions, 5 deletions

diff --git a/synapse/config/_util.py b/synapse/config/_util.py
index cd31b1c3c9..c74969a977 100644
--- a/synapse/config/_util.py
+++ b/synapse/config/_util.py
@@ -12,7 +12,7 @@
 # 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.
-from typing import Any, List
+from typing import Any, Iterable
 
 import jsonschema
 
@@ -20,7 +20,9 @@ from synapse.config._base import ConfigError
 from synapse.types import JsonDict
 
 
-def validate_config(json_schema: JsonDict, config: Any, config_path: List[str]) -> None:
+def validate_config(
+    json_schema: JsonDict, config: Any, config_path: Iterable[str]
+) -> None:
     """Validates a config setting against a JsonSchema definition
 
     This can be used to validate a section of the config file against a schema
diff --git a/synapse/config/federation.py b/synapse/config/federation.py
index 2c77d8f85b..ffd8fca54e 100644
--- a/synapse/config/federation.py
+++ b/synapse/config/federation.py
@@ -17,7 +17,8 @@ from typing import Optional
 
 from netaddr import IPSet
 
-from ._base import Config, ConfigError
+from synapse.config._base import Config, ConfigError
+from synapse.config._util import validate_config
 
 
 class FederationConfig(Config):
@@ -52,8 +53,18 @@ class FederationConfig(Config):
                 "Invalid range(s) provided in federation_ip_range_blacklist: %s" % e
             )
 
+        federation_metrics_domains = config.get("federation_metrics_domains") or []
+        validate_config(
+            _METRICS_FOR_DOMAINS_SCHEMA,
+            federation_metrics_domains,
+            ("federation_metrics_domains",),
+        )
+        self.federation_metrics_domains = set(federation_metrics_domains)
+
     def generate_config_section(self, config_dir_path, server_name, **kwargs):
         return """\
+        ## Federation ##
+
         # Restrict federation to the following whitelist of domains.
         # N.B. we recommend also firewalling your federation listener to limit
         # inbound federation traffic as early as possible, rather than relying
@@ -85,4 +96,18 @@ class FederationConfig(Config):
           - '::1/128'
           - 'fe80::/64'
           - 'fc00::/7'
+
+        # Report prometheus metrics on the age of PDUs being sent to and received from
+        # the following domains. This can be used to give an idea of "delay" on inbound
+        # and outbound federation, though be aware that any delay can be due to problems
+        # at either end or with the intermediate network.
+        #
+        # By default, no domains are monitored in this way.
+        #
+        #federation_metrics_domains:
+        #  - matrix.org
+        #  - example.com
         """
+
+
+_METRICS_FOR_DOMAINS_SCHEMA = {"type": "array", "items": {"type": "string"}}
diff --git a/synapse/config/homeserver.py b/synapse/config/homeserver.py
index 556e291495..be65554524 100644
--- a/synapse/config/homeserver.py
+++ b/synapse/config/homeserver.py
@@ -92,5 +92,4 @@ class HomeServerConfig(RootConfig):
         TracerConfig,
         WorkerConfig,
         RedisConfig,
-        FederationConfig,
     ]
diff --git a/synapse/config/tls.py b/synapse/config/tls.py
index e368ea564d..9ddb8b546b 100644
--- a/synapse/config/tls.py
+++ b/synapse/config/tls.py
@@ -471,7 +471,6 @@ class TlsConfig(Config):
         # or by checking matrix.org/federationtester/api/report?server_name=$host
         #
         #tls_fingerprints: [{"sha256": "<base64_encoded_sha256_fingerprint>"}]
-
         """
             # Lowercase the string representation of boolean values
             % {