From my own ten second scan of the exploit it appears that it sends an SQL query to a particular vulnerable version of the MS SQL server that causes a buffer overflow. To exploit it you need to.... send the SQL query to the server. One possible way to do this (as mentioned in the exploit) would be via SQL injection - like through a web application. Another way would be via a command line SQL client, which would require you to logon using a valid database userid that has the appropriate permissions to make the SQL query. If you understand the way that databases work this is pretty straightforward, so I suggest you do some research on this topic - maybe search the Internet on how to get command line SQL access to a MS SQL server.