fixed logged channel update

Author: daniel-sanche

google/cloud/bigtable/data/_async/client.py

@@ -368,8 +368,8 @@ async def _manage_channel(
             # prepare new channel for use
             old_channel = self.transport.grpc_channel
             new_channel = self.transport.create_channel()
+            new_channel._unary_unary_interceptors.append(self.transport._interceptor)
             await self._ping_and_warm_instances(channel=new_channel)
-            new_channel.unary_unary_interceptors.append(self.transport._interceptor)
             # cycle channel out of use, with long grace window before closure
             # TODO: refactor to avoid using internal references: https://github.com/googleapis/python-bigtable/issues/1094
             self.transport._grpc_channel = new_channel

google/cloud/bigtable/data/_sync_autogen/client.py

@@ -282,8 +282,10 @@ def _manage_channel(
             start_timestamp = time.monotonic()
             old_channel = self.transport.grpc_channel
             new_channel = self.transport.create_channel()
+            new_channel._unary_unary_interceptors.append(self.transport._interceptor)
             self._ping_and_warm_instances(channel=new_channel)
             self.transport._grpc_channel = new_channel
+            self.transport._logged_channel = new_channel
             self.transport._stubs = {}
             self.transport._prep_wrapped_messages(self.client_info)
             if grace_period: