Dimension data update in spark streaming without stopping streaming app

Question

Dimension data refreshes every hour. Data is huge so it is cached and used for streaming batches for that hour. As of now I have to kill the application and refresh dimension da