I have a mysql cluster setup and working. When connecting to it I specify one of the mysql nodes (mysqld) in my connection string. However, if this node goes down it doesn't automatically fail over to the other mysqld node (obviously). When someone creats an app to connect to the cluster do they simply specify each mysqld node in their connection string? In a MSSQL cluster you have a virtual IP you connect to so that instead of specifying each SQL node.
It's dependent on the connector that you're using. For example, if using Connector/J then you can provide multiple server addresses in your connect-string... http://dev.mysql.com/doc/connector-j/en/connector-j-reference-configuration-properties.html