- Question: When calculating latency time from host to destination, which values do we use in the traceroute result? Each hop has 3 time fields listed after it. I'm not sure which to use.
Answer: Traceroute lunches 3 probes (by default). Just pick the smallest one.
- Question: In Problem 3 (Prob. 5 on P. 67), the question asks to find the value of S that minimizes the delay of moving THE PACKET from Host A to Host B. If this is the case, we can simply use the smallest packet size which will result in shorter transmission time for moving A PACKET from Host A to Host B.
Answer: "THE PACKET" should be interpreted as "THE FILE". So you want to minimize the total delay of moving the file from Host A to Host B.
- Question: Does the sample frequency (44.1 kHz) in Problem 4 have anything to do with FDM (Frequency-Division Multiplexing)?
Answer: No. If CD is recorded with 22.05 kHz, you can put roughly 148 minutes worth of music in a 650 MB CD.
- Question: What number in the output of traceroute should we use as the latency?
Answer: Here is an example (traceroute to HBU - Houston Baptist University).
Bayou.UH.EDU> traceroute www.hbu.edu traceroute to www.hbu.edu (18.104.22.168), 30 hops max, 40 byte packets 1 caesar-atm12-00-1.gw.uh.edu (22.214.171.124) 1 ms 2 ms 0 ms 2 a1-3-1-0-3.a01.hstntx01.us.ra.verio.net (126.96.36.199) 2 ms 2 ms 2 ms 3 ge-0-1-0.r00.hstntx01.us.bb.verio.net (188.8.131.52) 2 ms 2 ms 2 ms 4 p4-4-0-0.r01.dllstx01.us.bb.verio.net (184.108.40.206) 7 ms 7 ms 7 ms 5 p4-2-0.r00.dllstx01.us.bb.verio.net (220.127.116.11) 7 ms 7 ms 8 ms 6 core4-serial0-0-0-0.Dallas.cw.net (18.104.22.168) 8 ms 9 ms 8 ms 7 corerouter2.Dallas.cw.net (22.214.171.124) 20 ms 10 ms 12 ms 8 acr2-loopback.Dallasdan.cw.net (126.96.36.199) 9 ms 10 ms 9 ms 9 aar1-loopback.Dallasdan.cw.net (188.8.131.52) 10 ms 10 ms 9 ms 10 internet-access-point-corp.Dallasdan.cw.net (184.108.40.206) 15 ms 16 ms 15 ms 11 gateway-005.Houston.iapc.net (220.127.116.11) 17 ms 22 ms 18 ms 12 18.104.22.168 (22.214.171.124) 22 ms 21 ms 20 ms 13 126.96.36.199 (188.8.131.52) 23 ms 22 ms 22 msThere are 3 numbers (23ms, 22ms, 22ms) in very last line (No. 13). They are round-trip time to the destination. So you should pick the shortest time and divide it by 2. In this case, the latency between bayou.uh.edu to www.hbu.edu is 11ms.
One interesting observation is that the route from UH to HBU actually goes through Dallas and come back to Houston, although HBU is located in the southwestern side of Houston.
- Question: Where can find out more about traceroute?
Answer: One good way is to read the man page of traceroute. It contains detail info about the command line parameters and the format of output.
- Question: In Problem 4, can I use MB/min as the unit of necessary bandwidth?
Answer: Well, this is part of my fault. Generally, when we discuss bandwidth in Computer Networks, the basic unit for bandwidth is bps (bits per second). Since in the third part of this problem ask you to compare with modem's bandwidth (56000 bps), you should convert all the bandwidth to bps or kbps.
There are two parts of questions in this assignment. The first part is written problems. For the second part, you have to search the answer from Internet and perform some experiments on the internet. This assignment weights 100 points.
The usual note:
- Homework Problem (2) in the textbook on page 66. (20 %)
- Homework Problem (3) in the textbook on page 66 ~ 67. (40 %)
- Homework Problem (5) in the textbook on page 67. (10 %)
- One 650 MB CD-ROM can hold 74 minutes worth of music (with 44.1 kHz sampling frequency). Assume that no data compression is done; this would in practice almost never be the case. What is the necessary bandwidth to transmit the music in real time?
MP3 (MPEG-1 Layer 3) typically has data reduction rate of 1:12 without losing the sound quality. What is the necessary bandwidth to transmit MP3 music in real time?
Can a regular modem connection (56000 bps) satisfy the bandwidth needed for MP3 music transmitted in real time?
- Internet standards are defined by Internet Engineering Task Force (IETF) and published as RFCs (Request For Comments).
(a) When is the first RFC published? What is the topic? Who submitted it?
(b) How many RFCs have been published? When is the latest RFC published? What is the topic? Who submitted it?
- This problem requires you to run a traceroute from your computer to some hosts and count number of hops required to reach the destination and also record the latency to reach the destination. You can run "traceroute using any of the following ways.
Here is what you need to do. First of all, pick three (3) of the following hosts and label them as H1, H2, and H3.
- from textbook companion website,
- run "tracert" on your Windows 95/98/NT/2000 machines, or
- if you have an account on bayou, you can run "traceroute" at prompt or /usr/sbin/traceroute if it can't be found.
- There is a web site called www.traceroute.org. This site provides links to hosts all over the world which provide interfaces to run traceroute from their local machines or routers. So you can run traceroute from Finland (EUNet) or Italy (IHnet).
Secondly, pick two hosts as routers from www.traceroute.org and label them as R1 and R2. Also, record their host names.
- www.rice.edu - Rice University
- www.cisco.com - Cisco Systems, Inc.
- www.msu.ru - Moscow State University, Russia
- www.tau.ac.il - Tel Aviv University, Israel
- www.nus.edu.sg - National University of Singapore, Singapore
Thirdly, run traceroute from source (R1 and R2) to destination (H1, H2, and H3). Record number of hops and latency between source and destination. You should have 6 values for number of hops and 6 values for latency.
Fourthly, pick a home site. You can use your own PC (running Linux/BSD/Windows 95/98/NT/2000), bayou, or any host from www.traceroute.org (other than R1 and R2). We will label it as HS. Record host name or IP address of HS.
Fifthly, run traceroute from source (HS) to destination (R1, R2, H1, H2, and H3). Record number of hops and latency between source and destination. You should have 5 values for number of hops and 5 values for latency.
Now draw a graph similar to this one. Replace H1, H2, H3, R1, R2, and HS with their host names or IP addresses. Label each link with number of hops and latency.
Finally, calculate the total number of hops and latency between home site (HS) and hosts (H1, H2, and H3). Also, calculate the total number of hops and latency from home site (HS) via routers (R1 or R2) to hosts (H1, H2, and H3). You will have a table like this.
Paths # of hops Latency (msec) HS-H1 + HS-H2 + HS-H3 HS-R1-H1 + HS-R1-H2 + HS-R1-H3 HS-R2-H1 + HS-R2-H2 + HS-R2-H3
Posted: September 6, 2000
Last modified: September 12, 2000.
tihuang at cs . uh . edu