Commit 03751704 authored by Chad Barb's avatar Chad Barb
Browse files

This is the first version of the new tutorial which I have written which should

incorporate some of the changes in the system since calfeld's last tutorial.
This isnt totally done yet.

Note that this is in the doc subdirectory.. It is probably a Good Thing to start
moving the documentation files onto the web server.
parent 10973d41
<title>Utah Network Testbed - Tutorial</title>
<h1>Utah Network Testbed - Tutorial</h1>
<h3>First of all</h3>
First of all, to make sure you have a working account
on the Utah Network Testbed, log in via the
<a href="">Testbed Web Interface.</a>
If you do not have an account, don't worry-- the login page contains
detailed information about obtaining and verifying an account.
<h3>Designing a Network Topology</h3>
The Utah Network Testbed's power lies in its ability to assume many
different topologies; the description of a such a topology is a necessary
part of an experiment.
The Utah Network Testbed uses the "NS" ("Network Simulator")
format to describe network topologies.
This is substantially the same format used by
<a href="">ns-2</a>.
Since the Utah Network Testbed interface doesn't interpret these
files directly with Tcl, some functionality may work differently than
you expect, or not at all. Also,
some testbed-specific commands have been added. These begin
with "#TB"; they will be ignored as comments
by ns utilities outside of the testbed.
In short, most NS files should work on both the Testbed and ns-2, most
of the time.
For those unfamiliar with the NS format, here is an example.
Let's say we are trying to create a test network which looks like
the following:
<br><img src="vlanpic.png"><br>
(A is connected to B, B to C, and B to D.)<br>
An NS file which would describe such connectivity is as follows:
# begin example.ns
# This is a simple ns script. Comments start with #.
set ns [new Simulator]
# First, define some nodes.
set A [$ns node]
set B [$ns node]
set C [$ns node]
set D [$ns node]
# Second, make some links between them, specifying
# bandwidth, latency, and queue type
$ns duplex-link $A $B 100Mb 10ms DropTail
$ns duplex-link $B $C 100Mb 10Ms DropTail
$ns duplex-link $B $D 100Mb 10Ms DropTail
# If you are running FreeBSD or Linux on any nodes
# you can set the IP addresses like this:
# #TB set-ip-interface
# For instance, the next line sets the IP address of B on the port going to C:
#TB set-ip-interface B C
# Note: The initial '#' is required to insure that 'ns'
# will ignore these testbed-only commands.
# The Testbed software will choose IP addresses for you, but you can choose
# them if you prefer. It is generally a good idea to have every link from
# any given node be on a different subnet, and to have both ends of a link
# on the same subnet.
$ns run
# end example.ns
<h3>Beginning the Experiment</h3>
After logging on and validating to the Testbed Web Interface,
choose "Begin an Experiment." Fill in the Name and Long Name fields.
In the "Your NS file" field, place the <i>local path</i> of a NS file
which you have created to describe your network topology. This file
will be uploaded through your browser when you choose "Submit."
After submission, the testbed interface will begin processing your
request. This could take a few minutes, depending on how many nodes you
are using and what other features (such as delay nodes and bandwidth limits) you are using. Once processing is done, the interface will present you with
a list of nodes allocated for the experiment.
If you use the above .NS file, you should get a set of allocated nodes similar
to this (physical machine names will likely be different.)
Node Mapping:
Virtual Physical
-------------------- --------------------
delay1 tbpc15
A tbpc16
delay2 tbpc21
B tbpc19
C tbpc11
D tbpc18
delay0 tbpc07
IP Addresses:
Node IFC Destination IP
-------------------- -------------------- --------------------
B eth0 C
B eth2 A
A eth0 B
C eth0 B
D eth0 B
B eth1 D
Notice that three "delay nodes" have been allocated to accomodate the
10ms delay time, as specified in the .NS file.
<h3>So now you have your nodes</h3>
/* login, powercycle, reconfigure(delay must be fbsd) */
<h4>Logging in</h4>
Once nodes have been allocated for an experiment, they are rebooted,
and all members of your Project are given login accounts.
Your username and password for testbed machines
will be the same as the one you use on the web interface.
Use ssh to log into "". If your machine, as configured, supports
ssh, you can "ssh <machinename>" into one of your machines
("tbpc15" is one such machine in the above example.)
If you are concerned about ssh traffic interfering with an experiment,
or if you have configured a machine to run an operating system which does not
support incoming ssh connections, an alternative is
"tip <machinename>", which communicates directly with a machine's
serial port. The standard FreeBSD and Linux installations listen for logins
on this port.
Once you are logged in, use "sudo" to run commands as root.
<h4>Powercycling Nodes</h4>
Obviously, machines may be rebooted or powered off using local commands on the
machine, if available. Also, all machines are connected to a power controller,
accessable through "". While logged into
ops, use the commands "power on <machinename>", "power off <machinename>", and
"power cycle <machinename>" to control power to your nodes.
<h4>Reconfiguring Node Operating Systems</h4>
Each PC on the test bed has three partitions. One contains Redhat Linux 6.2,
another FreeBSD 4.0, and yet
another can be loaded with any suitable operating system. Such an operating system must
(obviously) run on an Intel Pentium II processor, but does not need to offer any specific
network support, such as IP, nor does it have to support ssh for logins (since "tip" can be
used to communicate through a machine's serial port.) Any PC which has been automatically
configured as a delay node by the testbed control system must run FreeBSD to fufill that
function. A system which is not a delay node will use Redhat Linux 6.2 by default.
Nodes can be reconfigured via the "" web interface. Once logged in, and
assuming an experiment has been started in a project you belong to, go into the
"Experiment Information" section, then click on the Experiment ID (the short name)
you assigned to the project. This will show a page of information relating to that
experiment, including a list of the nodes allocated to that experiment. Click on
the icon next to the machine you want to reconfigure; this will bring up the Node Control
Center. In the box labeled "Def boot image" choose "RHL62-STD" or "FBSD40-STD", or an image
of your own devising. To creare such an "Alternative" image,
\ No newline at end of file
Supports Markdown
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment