Spark streaming DStream state on worker

[Renyi Xiong]

Hi,

I want to do temporal join operation on DStream across RDDs, my question
is: Are RDDs from same DStream always computed on same worker (except
failover) ?

thanks,
Renyi.