TCP, UDP & Ports (Module 2)

Question 1

TCP or UDP? a. used by TFTP

b. used by FTP
c. connection-oriented
d. connectionless
e. best-effort delivery
f. guaranteed delivery

A. TCP, UDP, TCP, UDP, UDP, TCP

B. UDP, TCP, TCP, UDP, UDP, TCP

C. TCP, UDP, TCP, UDP, UDP, TCP

D. UDP, TCP, UDP, TCP, TCP, UDP

Answer 1

B (UDP, TCP, TCP, UDP, UDP, TCP)

Explanation: TCP = Connection-oriented, guaranteed delivery via acknowledgement
UDP = Connectionless, best-effort delivery with no acknowledgement
FTP is an acknowledged transfer where TFTP is a best-effort transfer

Question 2

Which of the port numbers in the table below is not associated with the following protocols?

FTP (data)
DHCP (client)
SMTP
SSH
HTTPS
SNMP

A. 21
B. 68
C. 25
D. 161
E. 20

Answer 2

A (21)

Question 3

Which two of the following accurately describe the TCP session establishment process and the
termination process?

A. a and c
B. a and d
C. b and c
D. b and d

Answer 3

C (b and c)

Explanation: TCP communication is critical to network functioning. The TCP process to establish a
session is for the first host to send a SYN to initiate the dialog. The far end sends an ACK to
acknowledge the request. Then the first host sends a SYN-ACK to establish the session. Tearing down a
session is a 4-step process. When a host is done, it will send a FIN. The far end will reply to the FIN with
an ACK. If the far end is done too, it will respond additionally with a FIN-ACK stating that it wan

Question 4

How would a host know if the UDP datagram it sent was received by the far end host?

A. TCP ACK
B. UDP ACK
C. CRC
D. It wouldn’t
E. ICMP

Answer 4

D (It wouldn’t)

Explanation: UDP is connectionless and therefore a best-effort delivery. There is no
acknowledgement sent so there is no way of confirming if everything got there.

Question 5

Why does it makes sense that voice packets in an ip phone session use UDP instead of TCP protocol?

Answer 5

Explanation: Voice traffic is extremely sensitive to delay and packet loss. So, two things come into
play here. First, it takes time to acknowledge a TCP packet and voice does not have the time to spare.
Second, if TCP were in use and detected that a packet in a voice stream of packets did not make it to the
destination, sending the packet again would only cause more problems. The initial problem being that
some part of the conversation will be chopped off – which is annoying. Then imagine if 20 seconds later
that missing voice packet showed up – it would be in the wrong place at the wrong time and only cause
more poor voice quality. So, with voice, you use best effort and then configure QOS in the network.