WL#2801: Cluster should be able to notify mysqld process of graceful shutdown request

Currently when you shutdown the cluster gracefully the mysqld reports the following:
Description:
050902 17:13:02 [Note] NDB Binlog: cluster failure for atae/dcacache.
050902 17:13:02 [Note] NDB Binlog: cluster failure for
cluster_replication/apply_status.
Management server closed connection early. It is probably being shut down (or
has crashed). We will retry the connection.
Management server closed connection early. It is probably being shut down (or
has crashed). We will retry the connection.


we should create a way for the cluster to communicate to mysqld that a shutdown
of the cluster has been issued by the administrator and not the arbitrator. 

We should show messages that reflect a shutdown not a possible crash.