srv_resolver: add 15s timeout to DNS lookups

Author: amaanq

changelog.d/19026.misc

@@ -0,0 +1 @@
+Reduces the SRV DNS record lookup timeout to 15 seconds

synapse/http/federation/srv_resolver.py

@@ -26,6 +26,7 @@
 
 import attr
 
+from twisted.internet import defer
 from twisted.internet.error import ConnectError
 from twisted.names import client, dns
 from twisted.names.error import DNSNameError, DNSNotImplementedError, DomainError
@@ -145,7 +146,7 @@ async def resolve_service(self, service_name: bytes) -> List[Server]:
 
         try:
             answers, _, _ = await make_deferred_yieldable(
-                self._dns_client.lookupService(service_name)
+                self._dns_client.lookupService(service_name, timeout=(1, 1, 2, 4, 2))
             )
         except DNSNameError:
             # TODO: cache this. We can get the SOA out of the exception, and use
@@ -165,6 +166,10 @@ async def resolve_service(self, service_name: bytes) -> List[Server]:
                 return list(cache_entry)
             else:
                 raise e
+        except defer.TimeoutError as e:
+            raise defer.TimeoutError(
+                f"Could not resolve DNS for SRV record {service_name!r} due to timeout (50s total)"
+            ) from e
 
         if (
             len(answers) == 1

tests/http/federation/test_srv_resolver.py

@@ -68,7 +68,9 @@ def do_lookup() -> Generator["Deferred[object]", object, List[Server]]:
         test_d = do_lookup()
         self.assertNoResult(test_d)
 
-        dns_client_mock.lookupService.assert_called_once_with(service_name)
+        dns_client_mock.lookupService.assert_called_once_with(
+            service_name, timeout=(15,)
+        )
 
         result_deferred.callback(([answer_srv], None, None))
 
@@ -98,7 +100,9 @@ def test_from_cache_expired_and_dns_fail(
         servers: List[Server]
         servers = yield defer.ensureDeferred(resolver.resolve_service(service_name))  # type: ignore[assignment]
 
-        dns_client_mock.lookupService.assert_called_once_with(service_name)
+        dns_client_mock.lookupService.assert_called_once_with(
+            service_name, timeout=(15,)
+        )
 
         self.assertEqual(len(servers), 1)
         self.assertEqual(servers, cache[service_name])