I am not sure what you mean by streamRDD. In SparkStreaming the basic abstraction is DStream, which is basically a continuous stream of RDDs. Besides the usual operations (map, filter, reduce), if you want to do some custom RDD operation on every RDD in a DStream, there are a couple of options.
1. DStream.transform() This operation uses a RDD-to-RDD function to transform each RDD in the original DStream to new RDDs, thus creating a new DStream. The RDD-to-RDD function can use any of the RDD operations. Take a look at the Spark and Spark Streaming documentations.
2. DStream.foreach() This operation uses a RDD-to-Unit function to that is execute on each RDD in the original DStream. You can use this to do things like take the final processed data from a dstream and push the data to some external databased, filesystem, etc.