You should also be able to connect Splunk to Hive as shown here:
http://blogs.splunk.com/2015/02/25/splunk-db-connect-cloudera-hive-jdbc-connector/
And then, according to multiple pages, you can connect Hive to Cassandra:
http://planetcassandra.org/blog/hive-support-for-cassandra-cql3/
There is a Splunk ODBC driver, but it is for use with Microsoft Excel, Tableau Desktop, and MicroStrategy Analytics Desktop.
You can read about another method in this previous Answers posting and this interview/blog post.
Woodcock's question and suggestion seem worth considering, though!
I don't understand... You would like to take a DB that is designed for HUGE amounts of data and then not only send all of that data also into Splunk, but every change that is made to that data? If you are serious, contact your Splunk sales team, they will gladly send a team of PS engineers over to help you with POC!
The cassandra DB is currently part of the application. We are trying out splunk. If it works out, then Cassandra will be removed from the architecture.
On a temporary basis, we need a set up where we can run both.
DEFINITELY call Splunk. I am positive they will help you for free just for the bragging rights for your rip-and-replace use-case.