I am using cnosdb 3+2 cluster. When I kill a data node, the drop operation gives an error. Why? Is there anything else I need to do?

Question

In the process of using cnosdb, I started with 3+2 cluster, I created a database db1, and then I manually killed a cnosdb process, and then when I tried to drop database db1, the following error was reported:

422 Unprocessable Entity, details: {"error_code":"050019","error_message":"grpc client request error: grpc status: status: Unavailable, message: \"error trying to connect: tcp connect error: Connection refused (os error 111)\", details: [], metadata: MetadataMap { headers: {} }"}

Is my operation wrong? Or am I missing a step?