We see this quite a lot in our clusters and operation folks think some thing is wrong with cluster. This has no impact on the operation. Can we reduce the level to WARN or INFO?
ERROR com.datastax.driver.core.ControlConnection - No rpc_address found for host /a.b.c.d in host/ip:9042's peers system table. /a.b.c.d will be ignored.
Fixed for 2.0.10.
This is consistent with the level for the other situations where we ignore nodes (unsupported protocol version, wrong cluster name).