Support for CLIENT KILL - "Subcommand KILL" errors in cluster log

[admtech]

In the log file for our DragonflyDB cluster with Redis Sentinel there are many messages "Subcommand KILL not supported". The error usually appears when the sentinel process changes the master (e.g. by reboot or update).

Ubuntu 22.04.3 LTS DragonflyDB Version: 1.12.1 Redis Sentinel Version: 6.0.16 3 Server in Cluster (Master, Slave, Slave)

I20231115 15:10:02.825436  3599 server_family.cc:692] Load finished, num keys read: 275581
I20231115 15:10:11.234252  3595 server_family.cc:1773] Replicating 192.168.0.63:6379
E20231115 15:10:11.236002  3595 server_family.cc:1210] Subcommand KILL not supported
E20231115 15:10:11.236089  3595 server_family.cc:1210] Subcommand KILL not supported
I20231115 15:10:11.276466  3595 replica.cc:508] Started full sync with 192.168.0.63:6379
I20231115 15:10:12.518281  3595 replica.cc:527] full sync finished in 1.28 s
I20231115 15:10:12.518425  3595 replica.cc:603] Transitioned into stable sync