This is computer network class Select the statement that is correct. Demultiplexing is used to direct incoming packets to the correct process. Since UDP is a "connectionless" protocol, it is possible for requests to go to the wrong socket. If the network is having a lot of congestion, then the application can switch between TCP and UDP dynamically to adjust for this. Demultiplexing is needed only when you have multiple instances of the same application running on a host machine. Question 2 ( 2 points) Saved Why does DNS use UDP rather than TCP? Deverlopers assume that since the application requests cannot run until DNS happens first, then it is okay to make it wait for the services to provide for reliability. DNS requests are so frequent, and have to happen quickly, we don't want to add extra time for the services provided by TCP. To ensure that we get the mapping without any errors. DNS can tolerate some erros..