The second half of the lab will explore getting connected to the Internet via a local or national Internet Service Provider.
rintintin.Colorado.EDU 1 % traceroute xor.com traceroute to xor.com (192.225.33.1), 30 hops max, 40 byte packets 1 128.138.129.1 (128.138.129.1) 1.968 ms 0.931 ms 1.034 ms 2 cuatm-gw (128.138.138.10) 1.457 ms 1.527 ms 1.543 ms 3 205.169.250.5 (205.169.250.5) 10.875 ms 7.162 ms 6.488 ms 4 204.131.250.42 (204.131.250.42) 7.187 ms 8.868 ms 7.295 ms 5 csnden7001.csn.net (205.169.130.250) 12.191 ms 6.834 ms 9.390 ms 6 denver-cr1.bbnplanet.net (4.0.208.253) 7.798 ms 25.693 ms 28.259 ms 7 denver-cr2.bbnplanet.net (4.0.52.22) 10.137 ms 7.511 ms 7.786 ms 8 cointcorp.bbnplanet.net (4.0.212.250) 8.297 ms 11.751 ms 8.761 ms 9 gw16.boulder.co.coop.net (199.45.132.2) 9.834 ms 10.667 ms 9.426 ms 10 xor.com (192.225.33.1) 9.548 ms 9.478 ms 9.281 msThe number in parentheses is the IP address of the computer that the packet gets forwarded through. The numbers at the end are the times that it took for packets to get from me to each of those middle-man computers and back again. ms stands for milliseconds, so you can see that my packets get to xor.com pretty speedily. Here is a traceroute to the Computer Science Department at the University of California, Berkeley. The times are longer and you can see the route is not very direct geographically. Packets go to Denver, then to Willow Springs (in New Jersey or maybe near Washington DC) then Chicago (chi), then to Stockton, California (stk) and finally on to Berkeley.
rintintin.Colorado.EDU 2 % traceroute cs.berkeley.edu traceroute to cs.berkeley.edu (128.32.35.123), 30 hops max, 40 byte packets 1 128.138.129.1 (128.138.129.1) 2.554 ms 1.333 ms 2.464 ms 2 cuatm-gw (128.138.138.10) 2.680 ms 1.752 ms 1.895 ms 3 ncaratm-cuatm.co.westnet.net (204.131.62.14) 11.401 ms 261.111 ms 9.278 ms 4 border2-hssi1-0.Denver.mci.net (204.70.29.5) 148.422 ms 95.682 ms 6.962 ms 5 core1-fddi-1.Denver.mci.net (204.70.3.113) 4.106 ms 5.022 ms 4.475 ms 6 core4.WillowSprings.mci.net (204.70.4.73) 28.230 ms 28.970 ms 28.644 ms 7 sl-chi-15-H1/0-T3.sprintlink.net (206.157.77.90) 36.790 ms 35.915 ms 39.205 ms 8 sl-chi-1-P0/0/0-155M.sprintlink.net (144.232.0.142) 36.634 ms 37.127 ms 40.553 ms 9 sl-stk-20-H5/0-T3.sprintlink.net (144.228.10.54) 56.758 ms 55.098 ms 56.037 ms 10 sl-stk-16-F0/0.sprintlink.net (144.228.40.16) 56.300 ms 59.743 ms 59.714 ms 11 * sl-csuberk-1-H1/0-T3.sprintlink.net (144.228.146.50) 63.412 ms 62.538 ms 12 inr-666-dmz.berkeley.edu (198.128.16.21) 58.981 ms 57.452 ms 58.134 ms 13 inr-107-styx.Berkeley.EDU (128.32.2.1) 52.521 ms 55.276 ms 51.877 ms 14 inr-180.Berkeley.EDU (128.32.120.180) 86.798 ms 56.894 ms 61.456 ms 15 hofmann.CS.Berkeley.EDU (128.32.35.123) 62.485 ms 65.437 ms 64.039 msSometimes you will get a host that just doesn't answer in time for the traceroute program (he is impatient) and it will print a * instead of a time. Sometimes you will get a middle-man host (called a router) who is very busy and so the time to get to him and back is very long compared to the other times in the traceroute. If this occurs, try the ping program to see more detail about how that slow machine is doing. Here is an example of using ping to check on the CS machine at Berkeley. If i just do ping without any options it just tells me the Berkeley machine is alive:
rintintin.Colorado.EDU 16 % /usr/sbin/ping cs.berkeley.edu cs.berkeley.edu is aliveBut if I use ping -s then it sends a packet every few seconds and shows me both the time it took, and the sequence number of the packet. If the sequence numbers are in a row the machine at the other end is healthy; if they skip values that means that the packet with the missing value got lost or dropped because something on the way to the final host was too busy. Here is that same example, but with the -s option included:
rintintin.Colorado.EDU 15 % /usr/sbin/ping -s cs.berkeley.edu PING cs.berkeley.edu: 56 data bytes 64 bytes from hofmann.CS.Berkeley.EDU (128.32.35.123): icmp_seq=0. time=62. ms 64 bytes from hofmann.CS.Berkeley.EDU (128.32.35.123): icmp_seq=1. time=59. ms 64 bytes from hofmann.CS.Berkeley.EDU (128.32.35.123): icmp_seq=2. time=56. ms 64 bytes from hofmann.CS.Berkeley.EDU (128.32.35.123): icmp_seq=3. time=58. ms 64 bytes from hofmann.CS.Berkeley.EDU (128.32.35.123): icmp_seq=4. time=59. ms 64 bytes from hofmann.CS.Berkeley.EDU (128.32.35.123): icmp_seq=5. time=65. ms 64 bytes from hofmann.CS.Berkeley.EDU (128.32.35.123): ^C ----cs.berkeley.edu PING Statistics---- 7 packets transmitted, 7 packets received, 0% packet loss round-trip (ms) min/avg/max = 56/59/65
NOTE: to stop the ping -s I had to type control-c (listed in the output above as ^C). Hold down the control key, like you do the shift key and then type a c. Otherwise you will ping forever. ALSO NOTE: that ping typed without the /usr/sbin/ in front of it will say "Command not found:. This is because the ping command lives in the directory /usr/sbin and we mere mortals don't usually have that directory in our "path" (the place where using looks for things). If you want to add /usr/sbin to your path you would change the line in your file called .profile that looks like:
PATH=$HOME/bin:/usr/local/bin:/usr/ucb:/bin:/usr/binto look like:
PATH=$HOME/bin:/usr/local/bin:/usr/ucb:/bin:/usr/bin:/usr/sbinIf you decide to do this, to make this lab easier and to always have the ping command at your fingertips, and if you didn't add ~evi/bin to your path in lab 3 or 4 where we introduced the showgrades command, now would be a fine time to do it. Just add :~evi/bin to the end of that line.
Use traceroute and ping to find some busy hosts or
convoluted routes. I just tried the IRS's computer that has all
the tax forms on it and after about 12 hops it said !X which
according to the man page for traceroute, means that access is
administratively prohibited. Hmmm. The address i used was
irs.ustreas.gov. List the hosts you targetted, how many hops it
took to reach them and describe either the wierd geographical
route or the bottleneck machine. If you find a bottleneck,
use ping to explore which machine along the path is the baddie.
3. The finger command allows you to find who is logged into a computer, when they came to work, how long they have been idle (might be days if they forgot to logout). It has two forms, one where you tell it which user you want it to look up and theo other were you just tell it the machine and it gives you all the users. For example, finger evi@cs gives:
% finger evi@cs [cs.Colorado.EDU] <This is a lot of information. Some of it, such as the office data and office hours, I put in my file called .plan in my home directory. The information at the bottom show me logged on 4 times, one for each of the windows on my screen. It does tattle that I didn't make it in until 9:35 this morning and that two of my windows have been idle a long while. And it lies and says I have no unread mail (because my mail reader moves the mail out of the public area and into my home directory as soon as the mail arrives, so it always looks like I am caught up on mail). Use finger to check on your TA, to check on yourself, and to check on another faculty member you have for a class this semester. If you don't have a .plan file make one with the pico program. List something from your own "finger" info, from your TAs, and from one of your friends at another University.> To access the campus directory service, telnet cs.colorado.edu and login as 'da'. On-campus may use the 'da' or '411' commands. Fingering evi@rupertsberg.cs.colorado.edu ------------------------------------------ [rupertsberg.cs.colorado.edu] Login name: evi In real life: Evi Nemeth Directory: /home/columbine/evi Shell: /bin/tcsh On since Apr 16 09:34:46 on console No unread mail Plan: Evi Nemeth, ECOT 824 (used to be 8-3) 492-5385 office at CU 440-6093 x29 office at XOR 443-9832 home, OK till midnight spring '97: teaching csci 4830-010, "the internet for everyone" monday/wednesday at 4pm in duane g030, labs at various other times in eccr 244. out of town: feb 24 - mar 1 nluug conference, antwerpen mar 7 - 8 kentucky education meeting, louisville mar 11 uniforum conference, san francisco apr 6 - 11 ietf, memphis apr 24 - 25 sans conference, baltimore may 11 - 25 vacation, colorado river thru the grand canyon office hours: tues, wed noon-2pm and by appt. note: i am often in my office in the evening, feel free to stop by. also, reaching me by email usually works better than by phone. Login name: evi In real life: Evi Nemeth Directory: /home/columbine/evi Shell: /bin/tcsh On since Apr 16 09:35:13 on ttyp0 from :0.0 1 hour 48 minutes Idle Time Login name: evi In real life: Evi Nemeth Directory: /home/columbine/evi Shell: /bin/tcsh On since Apr 16 09:35:15 on ttyp2 from :0.0 Login name: evi In real life: Evi Nemeth Directory: /home/columbine/evi Shell: /bin/tcsh On since Apr 16 09:35:17 on ttyp3 from :0.0 1 hour 49 minutes Idle Time Login name: evi In real life: Evi Nemeth Directory: /home/columbine/evi Shell: /bin/tcsh On since Apr 16 09:35:18 on ttyp4 from :0.0
4. finger can also be used with just a machine name; finger @machinename will show you all users logged in there. For example, finger @ucsu would show you everyone who is currently logged in there. Try it. How many students are logged into it now? And to ucsub? (Don't count them, use the UNIX tool wc to do it for you.) When i just tried it I got
rintintin.Colorado.EDU 29 % finger @ucsu [ucsu.Colorado.EDU] Login Name TTY Idle When Office rc2 ??? dukeb duke ben pt 3d Sat 01:28 hansens HANSEN SALLY LORRAI *pt 1 Wed 11:42 girmusor GIRMUS-ORFORD ANNET *pt 1d Mon 14:59 goodrics goodrich stephanie *pt Wed 11:31and about 300 other students listed. The number toward the right shows how long they have been idle (3d, 1, 1d, and blank means 3 days for ben duke, 1 minute for sally hansen, 1 day for annet, and no idle time for stephanie). The day of the week and time are when they logged in. So we can figure out that ben logged in on Saturday in the wee hours but must have used the computer after that since it's more than 3 days since Saturday or really late Friday nite.
talk student-name@yoyo.cc.monash.edu.auTo exit talk you must type control-c. "talk" to someone local and to someone far away. Look at idle times to see who is really there. Who did you talk to?
6. Our final Internet exploration chore will be to run a couple of commands that give a totally different kind of information. Try each of these and tell us briefly what they said.
Providers you checked (must be at least 3) and how you found them:
Provider with the best service for your needs:
Provider with the cheapest prices for services you need:
8. Compare and contrast the 3 local providers: coop.net, sni.net, and rmi.net
based on information you can glean from their web pages. Present your data
in outline or tabular form so we can grade it more easily.
9. Use the UNIX tools (traceroute and ping for example) of the first
section of the lab to attempt to evaluate the quality of the ISP
that you are selecting from for your business. What is the round-trip time
to them from rintintin.