Real time process or thread priority is used to manage real time device operations, mostly in kernel mode. For example mouse or keyboard inputs may be set on real time as code for such device extends Real time priority class. Setting priority at real-time may be error prone, hence need expert knowledge, without which system operation may suffer. So, pre-defined priority setting is recomonded and preferred.
I've tried to make a general question in some sense, as I'm looking forward to gain a broader understanding of Windows scheduling engine.
However, my first thoughs about actually setting a process to Real-Time priority came when I experienced some issues with virtualized enviroments. In other words, it is about if whether setting a MS Windows proccess running a Virtual Machine Hypervisor would make the virtualized OS/VM more responsive.
I still have some doubts regarding the VM/Virtualization case, as it is stated by Microsoft (https://msdn.microsoft.com/en-us/library/windows/desktop/ms685100(v=vs.85).aspx) that a RT process would preempt even some I/O Interruptions/Services, like mouse, keyboard, disk ...
Therefoe, I guess an I/O intensive proccess running a Virtual Machine in a RT priority would suffer some penalties in this sense.
So. Would it be possible to conclude on beforehand that any I/O originated from/to the Real-Time VMM process which in its turn is running an I/O intensive proccess on the Virtualized OS, would preempt other proccess or I/Os from the native host OS which have a lower priority ?
PS: Feel free to answer in Brazilian Portuguese also, if you'd like
I would not mix realtime with quick processing. By defiition real time enables you only getting a guaranteed answering time. A guaranteed answer every 10 second would be also a real time process. So real time enables you that something is definitely happening also if the computer is very busy. So real time is something very usefull for data acquisition processes or to handle break conditions in a certain time window.