Out: Friday, January 6
Due: Sunday, January 15
Teams: No, do this individually
Read chapter 1 of Peterson text.
Please answer the following questions from the text:
Textbook Fifth Edition
Chapter 1: problems 3, 13, 16, 19, 20, 21, 22, 24, 26
- For #19.b even though it references 16.b it does ask about adding a single switch.
- For #21.b "latency" means the propagation delay.
- For #26.d you can assume that the image is a Binary Image.
Packets contain a time-to-live field (referred to as TTL). The value
in this field is decremented by one when it passes through a router
and the entire packet is dropped by the router if the counter reaches
zero. This mechanism exists in order to prevent packets from being
caught in persistent routing loops in the Internet and unnecessarily
consuming network bandwidth without being delivered to the appropriate
destination. (We will cover material later in the course as to why
these persistent routing loops appear in today's Internet.) The
source typically sets an initial value for the TTL field, which then
serves as the upper bound on the number of hops traversed by the
packet. If the packet is dropped when the counter reaches zero, a
diagnostic packet is sent to the source from the router that drops the
packet, and this diagnostic packet contains the identity of the router
at which the packet is dropped. The identity is the IP address of the
router, and the traceroute tool performs a reverse DNS lookup (i.e.,
translates the IP address to a human readable name) and reports both
the IP address and the DNS name to the user. (We will cover name
lookups and DNS later in the course.)
Traceroute is a tool devised by Van Jacobson that uses this feature in
order to measure the path traversed by a packet from the source to the
destination. (See Sec. 5.6 of the Computer Networks textbook or
search online for material describing the operation of traceroute.)
The tool sends a sequence of packets with increasing TTL values --
first with a TTL value of 1, followed by TTL value of 2, and so on --
and it prints out the identity of the router which is a certain number
of hops from the source when it receives the diagnostic message
indicating that the packet is dropped at a certain router. This tool
also provides additional information such as: (a) it measures and
reports the round trip latency from the source to the intermediate
router (typically in milliseconds), (b) it issues multiple probes per
hop so as to be robust to packet loss and if one of the probes is
lost, it reports that with a "*" (unresponsive router hops also result
in the same print message), and (c) if the different probes to the
same hop reach different routers (e.g., due to path changes or load
balancing that is taking place inside the network), it reports the
identity of all of the routers reached by the probes. The tool stops
when the destination is reached or if the number of hops exceeds a
default value (typically 30).
In this exercise, you will use the traceroute tool and learn how to
interpret the results reported by it.
Perform a traceroute to www.facebook.com. You can use the
tool from the command line (within the Terminal application) on
Linux/Mac OS X by simply running "traceroute www.facebook.com"
and on Windows machines by issuing the following command from the
command prompt: "tracert www.facebook.com". Answer the following
questions.Include screenshots of your terminal in the answers.
What is the IP address associated with "www.facebook.com"
(for example, "18.104.22.168")?
What is the IP address and DNS name of the router that is just
one hop before the destination?
How many hops did it take for the traceroute to reach from your
machine to the destination?
Which link incurs the longest latency (or delay) from the source
to the destination?
Perform a traceroute to www.google.com. Also perform a traceroute
to www.google.co.kr and www.google.co.in
What is the IP address associated with "www.google.com"?
Where do you think that "www.google.com", "www.google.co.in", and
"www.google.co.kr" are physically located?
Do you observe any unresponsive router hops (i.e., routers that
don't send a diagnostic message when the packet is dropped)?
Are the round trip latencies observed to intermediate hops
increasing uniformly or are there abrupt jumps? If there are abrupt
jumps in measured latencies, can you explain why they occur?
www.traceroute.org provides a list of publicly available traceroute
servers. Perform a traceroute to your computer from one of these
servers. Also do a traceroute to that remote server from your
computer. Are the same routers traversed in both directions?
Turn-in (online) a
.pdf file with your answers.