I use the authentication and authorization in my single node Cassandra setup, But I frequently get the following error in Cassandra server logs,
ERROR [Share
This message is really a signal of something wrong with your setup - machines are overloaded, or something like.
Instead of disabling these settings completely (changing password or chaning role will require restarting of the nodes), I would suggest to do following instead:
roles_validity_in_ms
, permissions_validity_in_ms
& credentials_validity_in_ms
to some quite high value, something like month;roles_update_interval_in_ms
, credentials_update_interval_in_ms
& permissions_update_interval_in_ms
to some value, like a minuteIt also makes sense to tune permissions_cache_max_entries
if you have big number of users & tables.