Hi Paul, thanks for the reply.
Answering to your questions, the transaction log of tempdb is set on a different physical device as the tempdb data and both are set to autogrow. Only the tempdb transaction log disk filled all the available space. I am not sure what is really happening and I'm not comfortable in adding more space to the disk without knowing what's going on. I'm already monitoring through perfmon all the version stored in tempdb too.
What might be your suggestion to this case? If it's a long running transaction that is causing this issue, what's the best approach to solve this?