I need some more information about the RR scheduling algorithm.

If I have a bunch of processes that will take some cpu time and IO
time.
RR defines a certain timeslice (say 10 ms). (which is used for cpu
timing?)

if  the cpu finishes with one process, which takes for example 9 ms,
and no overhead, then that process is send from the ready queue to the
ioqueue.
At 9ms, the cpu will take the next process from the readyqueue, go for
10 ms, etc

what happens in the meantime with the ioqueue, when are those io jobs
going to be done?
Will the program run cpu processes, and also io processes
simultaniously?

thank you for any help/ideas.