If this is Tableau tech support saying this, I would guess they are right.
There is an ODBC driver for iSeries. Would not a generic connection using that driver work?
Also, depending on how fresh you need the data to be for reporting purposes, you could also use the iSeries transfer utility. It can quickly pull records resulting from a query and dump them to a flat file that Tableau could consume. You could schedule a data dump based on how often you need refreshed data (and how often it is feasible to do this based on the volume of data).
Big issue is that the flat file is over 3 million records-for 2 divisions. Once we add the rest of our divisions, this number will grow exponentially. We need daily refresh on the data, so this is causing significant heartburn.
Also, the generic ODBC pulls we've tried appear to have join issues to other data sources (e.g. iSeries query dump into static Excel tables) that we've not been able to overcome yet.
We'll continue to whack away at it.
I can definitely sympathize with the heartburn! We worked on an ETL project for a client where we tried using the ODBC driver to pull tens of millions of records, but the performance wasn't reasonable for a nightly load. Our second iteration was to use the iSeries transfer utility to pull records into flat files as a preliminary staging. That option worked fairly well as a nightly process. Finally, the client decided to use a change data capture tool for pulling data.
There's probably not an easy answer, but hopefully you'll find something that works.
I found this thread searching for something similar and i would like to point you to this link on IBM site
ODBC diagnostic and performance tools
Hope this helps.
I'm interested to learn more about the change data capture tool that you implemented with iSeries. Was this done using transaction logs? How did this solution work out?