代写COMP304 COMP304-23A Routing with OSPF

- 首页 >> CS

THE UNIVERSITY OF WAIKATO
Department of Computer Science
COMP304-23A — Advanced Networking & Cyber Security
Assignment 1 - Routing with OSPF
Version of February 13, 2023
1 Introduction
The goals of this assignment are:
• to improve your understanding of OSPF by using it in a network
• become familiar with the way routers are configured by using an interface
similar to that of a typical router
In this assignment, you will configure the internal routing for your network in
the mini-Internet. The next assignment will follow on from this assignment
and you will connect your network with the network’s of all the students in
your class by configuring external routing. By the end, as a class, you will
have created your own version of the Internet using the real protocols actually
used today on the Internet.
For this assignment, you will configure virtual software routers rather than
physical routing hardware. These routers are running FRRouting. FRRouting
(FRR)1
is a popular routing protocol suite. FRRouting speaks many routing
protocols including OSPF and BGP and installs the routing decisions into the
Linux kernel. It has a management console interface that is similar to Cisco’s
IOS commands. FRRouting is a “real” technology used in real networks (not
just for teaching).
Note: Unlike the labs, this assignment is an individual. Work on your own.
You may discuss the assignment in general terms but do not look at anyone
else’s configuration or show them yours. If you need help, ask the lecturer in
person or by email, or post on Moodle. Please do not post configurations on
Moodle.
1.1 The mini-Internet
You have already accessed mini-Internet for labs, but the information is re peated here. The mini-Internet is a teaching project developed by ETH
Zurich2
. As a student, you are each given your own Autonomous System
(AS) to manage and configure.
1https://frrouting.org/
2The mini-Internet project https://github.com/nsg-ethz/mini_internet_project
1
The mini-Internet gives you access to several Docker containers. These Docker
containers behave like real hosts, switches, and routers; with the main differ ence being that you cannot reboot or shut them down. All of these networks
run on a single server, mini.cms.waikato.ac.nz. In the second assignment,
you will connect your AS with other students in your class.
1.2 Accessing the mini-Internet
You can access the mini-Internet using the secure shell (ssh) command, you
can access it either from home or the lab. One host within your network is
pre-configured with a secure shell server which is exposed on a unique port
on mini.cms.waikato.ac.nz. Your unique port number is 52000 added to
your AS number. You will receive an individual email with your AS number
and password; it is your responsibility to keep it confidential. Please do not
change this password. If you want to simplify access, please set up an SSH key
instead, details of which are included in the first lab.
The command to connect to the mini-Internet is:
ssh root@mini.cms.waikato.ac.nz -p


1.3 Accessing Routers, Hosts and Switches
Using the ./goto.sh script in the proxy container you can access any router,
host, or switch you wish. You can always see the container you are logged
into by looking at the hostname on the left of your terminal. To exit any
container back to the proxy you can press ctrl+d or type exit. You can log
in to multiple hosts or routers at once, or even the same one twice.
Accessing routers. Access a router using ./goto.sh router.
For example to access the router HAML:
root@gXX-proxy:~# ./goto.sh HAML router
Now you are in the FRRouting CLI on the router HAML. See 2.1 for details
on how to use this interface.
Accessing hosts. Each location shown in Figure 1 has a host attached di rectly to the router. To access the host attached to a router use the command
./goto.sh host. For example to access the NEWY host use:
root@gXX-proxy:~# ./goto.sh NEWY host
Recall for the lab tasks, you could access hosts in the UNIV network as follows:
root@gXX-proxy:~# ./goto.sh UNIV west-1
2
Now you are in a standard Linux shell, where you can run commands like ip
just as you have done in the lab exercises.
Accessing switches. Accessing switches is not required for this course.
However, for completeness, use ./goto.sh UNIV to access a
switch. For example to access the EAST switch use:
root@gXX-proxy:~# ./goto.sh UNIV EAST
Now you are in a standard Linux shell, the switches are running the Open
vSwitch software.
1.4 Backing Up Your Configuration
You can create a backup of the running configuration on your routers using
the ./save_configs.sh command on the proxy host. ./save_configs.sh
will create a folder configs [date] [time] and a zipped version of that folder.
You should keep regular backups of your configuration. You will also need to
submit this zip file as the final version of your configuration.
You can copy this off the proxy host using the scp command. For example on
your home or lab machine run:
scp -P \
root@mini.cms.waikato.ac.nz:.zip ./dst/path/
Note: you supply the port number to scp using the capital ‘-P’ option, rather
than the lowercase ‘-p’ like with ssh. In the command above ‘\’ splits this
long line across two, a terminal will remove the ‘\’ and run the command as
if it was on a single line.
1.5 Mini-Internet Help
If you have problems, please email the lecturer as soon as possible, if you
have any issues accessing your containers or other problems. Also, check the
Assignment section on Moodle for FAQs about this assignment.
1.6 Network Topology
Figure 1 shows your layer 3 network, which your task is to configure in this
assignment. All up you have 8 routers to configure with OSPF. Each router
has a host attached directly to it. In this assignment, you will configure these 8
routers using OSPF to establish full network connectivity between all routers
and hosts. Additionally, there is a DNS server (attached to LOND), this
provides you with forward and reverse DNS name resolution for the links and
3
MEASUREMENT
X.0.199.2/24
DNS
198.0.0.100/24
6
BOST
5
NEWY
7
ATLA
ZURI
8 TRGA
4
3
PARI
HAML
2
Provider 1 (link 2)
Provider 1 (link 1)
Provider 2
Customer 1 (link 2)
Customer 1 (link 1)
Customer 2
IXP
Peer
1
LOND
You have been assigned the network prefix:
X.0.0.0/8
Where X is your AS number and Y is the
location number, for example, NEWY is 5
Router loopback address: X.[150+Y].0.1/32
Each location has a host attached (not shown)
Host network: X.[100+Y].0.0/24
--> host: X.[100+Y].0.1
--> router: X.[100+Y].0.2
Figure 1: The internal layout and address plan of your AS. You have been
assigned an entire /8 to number your network. X.0.0.0/8 where X is the
network number given to you in your email. Each router shown has a host
attached. The addresses and subnets which you should configure for each
interface are shown. In this assignment, you should not configure the external
links shown in blue, leave these interfaces unconfigured.
hosts in your network (based on the numbering shown). Your hosts are pre configured to use this DNS server for commands including traceroute. DNS
will start working once you configure OSPF to share routes and a default route
on each host.
In addition, there is a MEASUREMENT host attached to HAML. You will
use this in the next assignment. This diagram also includes external network
links — such as customers, peers, and the IXP — which you will also use in
the next assignment. For now, you don’t need to worry about them. Once
you configure OSPF in this assignment, it will share the measurement route
throughout the network, this is expected behaviour.
2 Using the Routers
2.1 FRRouting Command Line Interface
You will need to configure FRRouting via its command line interface. It
presents a command line configuration interface similar to that used by other
router vendors such as Cisco or Juniper. One of the goals of this assignment
is to become familiar with an interface like that of a physical router.
You’ll find this interface is similar to using the Linux commands. For example,
instead of using the Linux command:
host# ip addr add 10.0.1.2/24 dev port_BOST
In FRRouting you would run:
router# configure terminal
router(config)# interface port_BOST
router(config-if)# ip address 10.0.1.2/24
When you first open the FRR CLI you begin in enable mode. To make
any changes we first have to enter into configuration mode. This is what
the configure command does. Then we enter the interface that we want
to configure (interface port_BOST). And finally set the IP address. From
here we can type end to return to enable mode. If we wanted to remove
this IP address, we would re-run the command prefixed by no. For example:
router(config-if)# no ip address 10.0.1.2/24
You can get help in the FRRouting CLI at any point by using and ?.
Typing will show you command completion options. And typing ? will
show you the commands available along with help.
Additionally, when in enable mode, you can use show running-config to print
out your configuration at any time.
You will be able to find documentation for FRRouting and OSPF on the web
that will help guide you through configuring FRRouting.
The official FRRouting documentation is here, we are running version 7.5.1:
http://docs.frrouting.org/en/latest/. For a quick start guide, see con figuring interfaces: https://github.com/rsanger/mini_internet_project/
wiki/2.5.2-Configuring-router-interfaces. Configuring OSPF: https:
//github.com/rsanger/mini_internet_project/wiki/2.5.4-Configure-OSPF.
Note: your assignment is different to theirs and has some extra steps not cov ered in that guide.
Also, check the Assignment section on Moodle if you are having trouble.
5
3 Configuring the Networks
This assignment has the following layout. You first demonstrate your ability to
plan the addresses on a small network. You will then configure your network
using the addressing plan provided. You will configure OSPF, and then finally
configure better metrics on your links to reduce the latency of the paths OSPF
selects.
3.1 Keeping Notes
You should keep track of the commands that you use for each step in this
assignment. You’ll find you often will need to come back to them. Also, take
note of any issues you have and submit this along with your configuration. A
full report is not required, these notes do not have to be typed, submitting
photos of handwritten notes is absolutely fine.
Some steps require you to submit evidence; these are highlighted in bold text.
Make sure you include the required information in your submission.
You will find that you are required to enter similar configuration on all of the
routers. You may find it useful to copy-paste the same series of commands
with addresses etc. updated where appropriate.
3.2 Addressing a Small Network
5
NEWY
7
ATLA
ZURI
8
15 Hosts
19 Hosts
9 Hosts
Figure 2: A portion of Figure 1 for you to address.
The address plan provided for you to follow in this assignment does not make
efficient usage of the address space. However, planning the efficient usage of
IPv4 address space is an important skill to master.
6
1. Following the process described in the IP addressing lectures, determine
the smallest prefix size that you would need to address the network shown
in Figure 2. Remember to allocate subnets for loopback addresses, the
three links between the routers, and the switched networks to support
the number of hosts listed.
2. Once you have determined the smallest prefix size, assign addresses out
of the range X.0.0.0/ (where X is your assigned AS number) to
the subnets.
Include both a copy of your working and the network addresses you
have allocated for each subnet. You can submit this as an image.
3.3 Interface Configuration
The first step in configuring your network is assigning the correct IP addresses
to interfaces. Refer to Figure 1 for the numbering to use. Routers and hosts
have been pre-configured with interface names. The interfaces are as follows:
Router: Configure IP addresses of these interfaces
host connected directly to the host in the same location
lo the loopback interface
port_DEST the port connected to the DEST router, e.g. port_HAML
Router: Do not change the configuration of these interfaces
ext_AS_* ports connected to external ASes, e.g. ext_8_TRGA
ssh pre-configured, ssh communication channel
dns_AS LOND, pre-configured, connects to DNS server
measurement_AS HAML, pre-configured, connects to measurement host
matrix_AS PARI, pre-configured used for measurement
TRGA-L2 a link to the UNIV network NORTH switch
Hosts:
DESTrouter configure, linked to the external router
lo the loopback, don’t configure on hosts
ssh pre-configured, ssh channel, don’t change
To configure an address on a router you will need to use the appropriate
FRRouting command (see Section 2.1), whereas you configure an address on
a host using the ip command like in the lab. You do not need to make the
configuration persistent on the hosts.
On the hosts, you only need to configure the interface connected to the router.
Unlike routers, hosts are not configured with publicly reachable loopback ad dresses.
3. Configure all interfaces between adjacent hosts and routers based on the
numbering in Figure 1. Your hosts should be numbered X.[100+Y].0.1/24,
and routers X.[100+Y].0.2/24, where X is your AS number and Y is
the location number shown beside each router in Figure 1. For example,
the host interface on the NEWY router should be assigned the address
7
X.105.0.2/24. Once configured, verify each host can reach its respective
router using ping.
4. Configure all router loopback interfaces based on the numbering in Fig ure 1. A router’s loopback address should be numbered X.[150+Y].0.1/32.
For example, PARI would be given the loopback address of X.153.0.1/32.
5. Configure addresses on all the internal links shown in Figure 1. Ensure
that each router can ping its directly connected neighbours via the di rectly connected interfaces. At this stage, there are only routes for the
directly connected networks; non-adjacent interfaces will not be able to
communicate.
3.4 OSPF Configuration
6. The next step is to configure OSPF on all of your routers (LOND, HAML,
PARI, TRGA, NEWY, BOST, ATLA, ZURI). Use the OSPF backbone area
0.0.0.0. On each router you need to first set the router id used by
OSPF to the router’s loopback address. Then tell OSPF which networks
to “speak” OSPF on and tell it it can share (a.k.a. redistribute) routes
that are known because the router is directly connected to the subnet.
Refer to the documentation in Section 2.1 to figure out how to do this.
Once they are exchanging network state and routes then each router
should be able to send IP packets to all other routers’ interfaces including
non-local subnets and the router loopback addresses.
It is best to start configuring two neighbouring routers first. Once that is
working, repeat this step for all remaining routers.
7. Check that all the enterprise routers can ping and traceroute all other
routers.
8. Add a default route on the hosts to their corresponding router. Then
ensure that hosts can reach all other hosts on the network.
9. Check each router’s routing table, the OSPF database and the OSPF
routing table. Spend a bit of time making sure you understand the
relationship between these tables. You might also like to look at the rest
of the OSPF information the router has.
3.5 Ensure OSPF Packets are not Leaked
It is bad practice to “speak” OSPF on interfaces without an OSPF router
attached. Doing so leaves your network vulnerable to route hijacking; where a
malicious host could form an OSPF adjacency and inject their own routes. In
this step, you will verify that the hosts are not receiving OSPF packets.
10. Verify the hosts are not receiving OSPF packets. The simplest way to do
this is to capture traffic on the hosts using tcpdump. Alternatively, look
8
at the OSPF interface configuration on the router. If hosts are receiving
OSPF packets, make changes to your OSPF configuration to fix this.
3.6 OSPF Traffic Engineering
The next step is to look at changing link costs to modify OSPF’s routing
decisions.
11. Measure and record the latency of all of the links between the routers.
The latency on the host links is negligible. Optionally, you can choose
to factor in the bandwidth of these links by running iperf3 between the
hosts. Note: you can run ping from a router, but not iperf3; iperf3 is
an application and only runs in hosts.
Link Latency Bandwidth (Opt.) Cost
LOND (1) < − > BOST (6)
LOND (1) < − > NEWY (5)
LOND (1) < − > PARI (3)
LOND (1) < − > HAML (2)
NEWY (5) < − > BOST (6)
NEWY (5) < − > PARI (3)
NEWY (5) < − > ZURI (8)
NEWY (5) < − > ATLA (7)
ZURI (8) < − > ATLA (7)
ZURI (8) < − > TRGA (4)
ZURI (8) < − > PARI (3)
PARI (3) < − > TRGA (4)
PARI (3) < − > HAML (2)
It may also be useful to sketch the latency on the network diagram.
Include the latencies you measured and the new link costs that
plan in Step 13 in your submission
12. By default, all links will have the same default cost, so OSPF will cur rently pick the path with the fewest hops. You can find the default cost
by running show ip ospf interface. Find an example of a particularly
high latency path.
In your submission, include the output of traceroute of this
high-latency path
13. Plan new link costs that prefer low-latency paths and apply the OSPF
costs on your network links. You don’t have to apply costs to all links
for this step, e.g. you can increase the cost for high latency links while
leaving all other links at their default. In FRR, you configure this cost on
the interface, rather than from within the router ospf configuration
item. Note: you will need to apply the new cost to both ends of a link,
otherwise it will only apply to traffic in one direction.
14. Verify that this has corrected the high-latency path you identified above.
In your submission, include the new output of traceroute
9
3.7 Finishing up
15. Check the routing table on all of the routers, they all should include
routes to every host network, link, and router loopback. Check hosts
can ping each other and router loopbacks. You can also check https://
mini.cms.waikato.ac.nz/matrix/matrix.html, which shows whether
a network is reachable from another. Your network AS number should
show as being reachable from itself, i.e. a green square on the diagonal
line for your AS number. You’ll use this connectivity matrix further in
Assignment 2.
16. Have a go at answering the questions on the next page. These aren’t
worth any marks, but are instead to help ensure you understand the
concepts.
17. Check over your configuration and make sure it is tidy. Use the no
command to remove any unintended lines from your router configuration.
Use show running-config to check for anything odd you have left in
your configuration. Verify reachability between your hosts. Save a copy
of your configuration using the ./save_configs.sh script and download
the .zip archive using scp (see Section 1.4). You must submit this .zip
on Moodle, along with your readme, link costs and addressing plan (see
the checklist in Section 5.1).
10
4 Questions
Answer the following questions to ensure you understand what is happening.
You will need to run commands to answer the questions.
Note: You do not need to submit your answers to the questions unless you are
unsure about an answer and would like it checked. There are no marks allo cated to the questions and you will not lose marks for submitting an incorrect
answer.
1. On each network to which the routers are connected, identify the OSPF
Designated Router.
2. What routes derived from OSPF appear in each router’s forwarding ta ble?
3. Which of these, if any, are OSPF external routes?
4. (a) What are the two numbers in [110/20]? What is each used for?
(b) Which route or routes derived from OSPF have the highest metric?
(c) Why is that metric value the number that it is?
5. (a) What link-state advertisements (LSAs) appear in each router’s topol ogy database?
(b) Which router LSAs, if any, describe an Autonomous System Bound ary Router (ASBR)? Explain why this is the case.
6. Demonstrate the frequency with which keep-alive (HELLO) packets are
being sent and received across an interface within the internal network
(pick any pair of internal routers). This question is about what is actually
happening, not what is configured.
7. Disable one link on a two-hop path and show that the traffic will take an other path with a higher metric/cost (Use (config)# interface port_X
followed by (config-if)# shutdown to disable an interface and
(config-if)# no shutdown to bring it back up again.
11
5 Assessment
Your assignment must be submitted in electronic form using Moodle. Do not
submit anything physical. You may submit either a set of files or an archive
containing multiple files made using tar or zip. Please do not use a different
archive program (like 7z for example).
Your submission must include:
• Documentation of the numbering plan for your routers. Submit this as
a png or jpeg file called number-plan.png or number-plan.jpg.
• Full and unmodified router configurations for all routers saved using
./save_configs.sh as described in Section 1.4 above. Upload this
zipped file without modification.
• The latencies you observed originally in the network, and your planned
costs. And the outputs of traceroute from before and after showing the
link you have corrected. Name these files with the name of the step (i.e.
step-11, step-12 and step-14) and submit these as either txt, pdf, png,
or jpg format.
• A readme.txt file containing the step you reached and any problems you
encountered.
• Optionally, a notes.txt or notes.pdf file containing any notes you have
taken.
Part Marks
Demonstrate assigning addresses to a network 1
IPv4 addresses configured on interfaces 1
Hosts configured with default routes 1
OSPF up and making all subnets (including the loop back addresses) reachable.
4
No OSPF packets leaking to hosts 1
OSPF link costs assigned correctly 2
10
12
5.1 Final Checklist
Used the correct IP address range as described in Figure 1
Submitted, via Moodle, number-plan.png or number-plan.jpg
Submitted, via Moodle, your configuration from save_configs.sh
Submitted, via Moodle, your latency measurements and metric plan
step-11.txt or step-11.jpg
Check you have:
Submitted, via Moodle, your high-latency traceroute step-12.txt
Submitted, via Moodle, your new traceroute named step-14.txt
Submitted, via Moodle, readme.txt
Re-read the submission and assessment instructions above.
13
A Testing and Troubleshooting
A.1 Ping and Traceroute
ping and traceroute are often the starting point of debugging. For example,
use ping to answer the questions: “What can you reach and what can’t you?”
and “Can you ping to each interface along the path you expect the packets to
follow?”
The traceroute command is also very helpful. Remember, though, that
traceroute shows the forward path only and that a failure at a particular hop
may be because the reply packets can not get back to the source.
A.2 Can’t ping adjacent interfaces
• Check that the interfaces are in the same subnet and that this is different
to all other subnets on each router.
• Use show interface brief to check that you have assigned the correct
address to the correct interface.
• If possible, try using capturing packets from a host using tcpdump to see
if the packets are leaving on the interface you expect and arriving on the
router interface.
• If you still can not ping try deleting and re-typing your interface configu ration (using no remove a part of the configuration) and then re-enter it.
But first, save a copy of your configuration using ./save_configs.sh so
you can refer back to it if needed.
• Finally, check and recheck your addresses and interface connections. It’s
too easy to “not see” a reversed or changed digit when you’re expecting
to see the correct values!
A.3 Forwarding and Route Information
The command show ip route (from operational mode) shows the routers for warding table. This includes the routes that have been selected from the
various routing protocols, static routes and directly connected networks. Note
that the source of the route is included in the output (e.g. O for OSPF).
The command show ip route ospf shows the routes in the forwarding table
that have come from OSPF (but not all the route that OSPF knows about...
see the OSPF section below for that).
A.4 OSPF Information
show ip ospf prints information about the OSPF area.
14
show ip ospf neighbor shows what neighbours OSPF is communicating with.
It’s a good quick check if something seems not to be working with OSPF.
show ip ospf interface gives information about the instance of OSPF that
is communicating over the subnet attached to that interface. It includes, for
example, the number of neighbours found and the identity of the DR and BDR.
show ip ospf route and show ip ospf database show the information that
OSPF knows about the network, including all the prefixes that it’s learned.
A.5 OSPF not working
1. Check the forwarding table show ip route to check that it is, in fact,
OSPF that’s not working. You should see both one and two-hop routes
marked as coming from OSPF.
2. Check whether OSPF is seeing its neighbours (show ip ospf neighbor).
If not, check that you have a separate network clause for each network
you want OSPF to operate on.
3. If OSPF isn’t seeing its neighbours, but you have correct network clauses,
check basic connectivity between the routers.
4. If OSPF can see its neighbours but no routes are being shared, check that
you have told OSPF to share (redistribute) the appropriate routes
(e.g. for step 3.4 directly connected networks).
15

站长地图