Do you happen to have the possibility of running the log file through some sort of ETL to flatten the data? If so, that might be a good approach. Even if you could pull it as is into a table in a database it would open up the possibilities of using joins.
But there is another possibility. You could use Tableau's data blending. This approach would be to create a copy of your data connection and then blend it on itself using the "OrderID" as a linking field. If you are not familiar with data blending, there is a good walk-through here here: http://kb.tableausoftware.com/articles/knowledgebase/relate-summarized-data-60
I've attached a Tableau workbook using your data (and some additional duplicate records) to demonstrate how it might be accomplished. Hope it gets you going!