mat379.sp05.ExamAnswers

advertisement
Computer Networking (MAT 379)
Final Exam
Spring 2005
Name:_______________________________
1. (5 points) What is the binary equivalent of 192.1.3.27?
11000000.00000001.00000011.000011011
2. (10 points) Suppose a LAN uses Class C IP addresses.
i. Which of the 32 bits will denote the network portion of the address? first 24 bits
ii. Give the netmask in dotted decimal format. 255.255.255.0
iii. How many hosts can a network have? 28-2 = 254
3. (5 points) Mention three differences between the IPv4 and IPv6 datagram headers?
i. IPv6 has 128 bit fields for IP addresses.
ii. IPv6 has a fixed size header since are options located within the data section of the datagram.
iii. IPv6 provides a flow label field to identify packets that require special handling, e.g., audio and video
transmissions or traffic with a high priority.
4. (5 points) The Routing Information Protocol (RIP) uses UDP for its underlying transport
protocol. How does RIP overcome the unreliability of UDP?
A RIP router rebroadcasts its routes every 30 seconds.
5. (5 points) Briefly state the purpose of TCP flow control and explain how it works.
TCP flow control prevents a sender from sending data faster than a receiver can process the data. A TCP
sender can never send more unacknowledged data than will fit in the available space in the receiver’s
receive buffer.
6. (15 points) Consider a router that interconnects three subnets. Suppose all three subnets must
have the prefix 223.1.17/24. Suppose Subnet 1 is required to have 125 interfaces, and Subnets 2
and 3 each are required to have up to 60 interfaces. Provide three network addresses of the form
a.b.c.d/x that satisfy these constraints. Answer on back. Explain and show work. (Hint: consider
satisfying the constraints using the binary form of the address then convert to decimal.)
There are several possible ways to assign addresses that satisfy the above constraints. Here is one.
Since the first 24 bits are fixed, some of the last 8 bits must be used to identify the new networks.
0 _ _ _ _ _ _ For the first network, let the first bit of the last 8 bits denotes the network so the first 24 +
1 (25) bits indicate the whole network address. Since the last seven bits can be used for
hosts, this network can have 27-2 or 126 hosts.
1 0 _ _ _ _ _ For second network, let the first two bits of the last 8 indicate this network so 24 + 2 (26)
bits are needed for the network address.
1 1 _ _ _ _ _ In the third network, again let the first two bits of the last 8 indicate the network so 24 + 2
(26) bits are needed for the network address.
In the second and third networks, there are 6 bits available for hosts. Thus, each of these networks can
support 26-2 or 62 hosts. Converting the binary in the last octet to decimal, the networks can be specified
as follows:
223.1.17.0/25, 223.1.17.128/26, and 223.1.17.192/26
7. (20 points) Assume that the Distance-Vector Algorithm is used to find minimum cost routing
tables for the nodes in figure 1. Answer the following questions.
Figure 1
2
X
Y
7
1
Z
i. Give the initial routing tables at each node.
Node X
Destination Cost
Y
2
Z
7
Node Y
Destination Cost
X
2
Z
1
Node Z
Destination Cost
X
7
Y
1
ii. Suppose Y sends its initial routing table to Z. Give the updated routing table at node Z.
Node Z
Destination Cost
X
3
Y
1
iii. Give the optimal routing tables at nodes Y and X. These are the routing tables that appear
after the Distance Vector algorithm has converged.
Node Y
Destination Cost
X
2
Z
1
Node X
Destination Cost
Y
2
Z
3
iv. Suppose the cost from X to Y changes from 2 to 10. Give the Node Y optimal routing table
after the change.
Node Y
Destination Cost
X
8
Z
1
v. Explain how the optimal routing table at Y is established and indicate the number of messages
that would be required.
After the cost from Y to X over the direct route changes to 10, Y will receive a message from Z saying it
can get to X at cost 3. Y will then think it can get to X through Z at a cost of 4. It will broadcast to Z. Z
will think it can get to X for a cost of 5. This continues as follows: Z broadcasts, Y thinks YX cost 6;
Y broadcasts, Z thinks ZX cost 7; Z broadcasts, Y thinks Y X cost 8; Z broadcasts, Y thinks Y X
cost 8. It takes about six messages for the system to reach this steady state.
8. (15 points) Consider the link-state algorithm used to find the minimum path distances from
node A to each of the other nodes in the network in the following figure.
Complete the following table for each step of the algorithm until it terminates.
Step N
0
{A
1
{A,E}
D(B) p(B) D(C) p(C) D(D) p(C) D(E) p(E)
} 7,A
Infinity
Infinity
1,A
4,E
Infinity
3,E
2
{A,E,D}
6,D
3
{A,E,D,B}
5,B
4
{A,E,D,B,C}
In the table, N denotes the set of nodes processed up to the corresponding step, D(X) denotes the
current shortest path distance estimate from node X to A, and p(X) denotes the predecessor of
node X. Note that the algorithm may terminate in less than 4 steps.
9. (20 points) Consider the following plot of TCP windows size as a function of time.
Assume that TCP RENO is the implementation of TCP congestion control used.
i. Identify the intervals when TCP slow start is operating.
Interval 2-6 and 23-26
ii. Identify the intervals when congestion avoidance is operation.
Intervals 6-16 and 17-22
iii. How is segment loss detected after the 16th round?
A triple duplicate ACK
iv After the 22nd transmission round, is segment loss detected by a triple duplicate ACK or a
timeout.
A timeout
v. Give the value of Threshold at the following rounds:
1st round: 32 , 18th round: 21
, 24th round: 13
vi. Assuming a packet loss is detected after the 26th round by the receipt of a triple duplicate
ACK, what will be the values of the congestion-window and the size of Threshold?
congestion-window: 4
Threshold: 4
10. (10 points) Border Gateway Protocol
i. Mention and briefly describe two features of the Border Gateway Protocol (BGP) that make it
more suitable for routing in the global Internet than either the Routing Information Protocol
(RIP) or the Open Shortest Path First (OSPF) Protocol.
BGP allows policy routing. In other words, routers can choose routes based on factors other than the
shortest number of hops to a destination. For example, an ISP might want to limit the routes it uses to
only those ISPs with which it has contracts.
BGP scales better than RIP or OSPF. BGP routes according to autonomous systems, which are large
collections of networks under a single administrative authority.
ii. How does BGP avoid introducing routing loops into paths?
A BGP router sends the list of autonomous systems it traverses to reach a destination (AS-Path attribute).
If a receiving BGP router sees itself in the list, the router rejects this path.
Download