Hi team, we've noticed a massive amount of "finished call" logs in spicedb. Something like 400 events per second per spicedb pod. Is this something expected? We can just filter out the log, but the level is "info" so maybe we are doing something wrong, like not properly reusing the connection? How long are connections to spicedb expected to last? Do they have to be reused accross independent data requests?