Change the Default Log Level from Trace to Warn

[schromp]

For some reason the default log level in the crds is set to trace.
This lead to 44M (Raft) Log Lines being pushed to our monitoring stack in 15 minutes.

Standard practice is setting log level to warn, afaik.

clickhouse-operator/config/crd/bases/clickhouse.com_keeperclusters.yaml

Line 6108 in 203c865

default: trace

[ashishch432]

The operator level is also set to debug as default, which could be revisited.
On a related note to logging - when logToFile fails it constantly prints stderr write failures which can bulk up logging if that ever happens for any reason (temp storage running out etc.)