MXS-2223 Log a message when a slave is discriminated due to replication lag

Both the replication lag and the message printing state are saved in SERVER,
although the values are mostly used by readwritesplit. A log message is printed
both when a server goes over the limit and when it comes back below.
Because of concurrency issues, a message may be printed multiple times before
different threads detect the new message state.

Documentation updated to explain the change.

server/core/server.cc

@@ -183,6 +183,7 @@ SERVER* server_alloc(const char* name, MXS_CONFIG_PARAMETER* params)
     server->master_id = -1;
     server->master_err_is_logged = false;
     server->warn_ssl_not_enabled = true;
+    server->rlag_state = RLAG_NONE;
     server->disk_space_threshold = NULL;
 
     if (*monuser && *monpw)