Clean up ShardedWorkerHandlingConfig (#9466)

* Split ShardedWorkerHandlingConfig

This is so that we have a type level understanding of when it is safe to
call `get_instance(..)` (as opposed to `should_handle(..)`).

* Remove special cases in ShardedWorkerHandlingConfig.

`ShardedWorkerHandlingConfig` tried to handle the various different ways
it was possible to configure federation senders and pushers. This led to
special cases that weren't hit during testing.

To fix this the handling of the different cases is moved from there and
`generic_worker` into the worker config class. This allows us to have
the logic in one place and allows the rest of the code to ignore the
different cases.

synapse/config/_base.py

@@ -844,22 +844,23 @@ class ShardedWorkerHandlingConfig:
 
     def should_handle(self, instance_name: str, key: str) -> bool:
         """Whether this instance is responsible for handling the given key."""
-        # If multiple instances are not defined we always return true
-        if not self.instances or len(self.instances) == 1:
-            return True
+        # If no instances are defined we assume some other worker is handling
+        # this.
+        if not self.instances:
+            return False
 
-        return self.get_instance(key) == instance_name
+        return self._get_instance(key) == instance_name
 
-    def get_instance(self, key: str) -> str:
+    def _get_instance(self, key: str) -> str:
         """Get the instance responsible for handling the given key.
 
-        Note: For things like federation sending the config for which instance
-        is sending is known only to the sender instance if there is only one.
-        Therefore `should_handle` should be used where possible.
+        Note: For federation sending and pushers the config for which instance
+        is sending is known only to the sender instance, so we don't expose this
+        method by default.
         """
 
         if not self.instances:
-            return "master"
+            raise Exception("Unknown worker")
 
         if len(self.instances) == 1:
             return self.instances[0]
@@ -876,4 +877,21 @@ class ShardedWorkerHandlingConfig:
         return self.instances[remainder]
 
 
+@attr.s
+class RoutableShardedWorkerHandlingConfig(ShardedWorkerHandlingConfig):
+    """A version of `ShardedWorkerHandlingConfig` that is used for config
+    options where all instances know which instances are responsible for the
+    sharded work.
+    """
+
+    def __attrs_post_init__(self):
+        # We require that `self.instances` is non-empty.
+        if not self.instances:
+            raise Exception("Got empty list of instances for shard config")
+
+    def get_instance(self, key: str) -> str:
+        """Get the instance responsible for handling the given key."""
+        return self._get_instance(key)
+
+
 __all__ = ["Config", "RootConfig", "ShardedWorkerHandlingConfig"]