Working of Data MDT has already been defined in recent post.
This post will actually cover the implementation of data mdt in network. Data MDT is configured under vrf. You cannot configured more than 256 addresses per vpn or vrf. One thing keep in mind during its implementation that data mdt range should be different from default mdt range while using auto-rp or bsr. If you are using the SSM then the data and default mdt can be same. This is the design problem with sparse mode.

Ip vrf Shivlu
mdt data 239.1.1.0 0.0.0.255 threshold 1

Threshold means that whenever the multicast stream will trigger more than 1 Kbps a data mdt will fire. You can say multicast tree will be constructed on demand.
Note:- Data MDT will be created only by the PE that has the source connected to it.
As i have already told we are having limited number of data mdts for particular vpn so we can re-use the addresses with the help of ref-count. Now the question comes which number will be used after 256. The data mdt addresses which are used least that number will be reused.

How to check the data mdt usage
“ Show ip pim vrf Shivlu mdt send “ command will help you to check the data mdt operation. In the output you can see the ref_count.
How to check data mdt is used for sending and receiving
Run the command “show ip mroute vrf Shivlu “ and in the output if you are getting the Y/y flag it means data mdt is working fine.