Skip to content

Latest commit

 

History

History
218 lines (177 loc) · 8.18 KB

week-08-network.md

File metadata and controls

218 lines (177 loc) · 8.18 KB
Raw

General Questions

  1. Explain the purpose of the time-to-live field in an IPv4 datagram header.

  2. If a single faulty router on a network incremented TTL instead of decrementing it, how bad would it be for the network as a whole?

  3. Why does including both TTL and a checksum make IPv4 less efficient?

  4. There are two ways that an IPv4 receiver can tell that a datagram it receives is a fragment. What are they? Must both indications be present in every fragment?

  5. Is it possible for an IP datagram successfully traverse several links before being fragmented, or must fragmentation occur at the first link? If it is possible, explain why or give an example of how this could happen. If it is not possible, explain why not.

  6. Two movers are trying to get a treadmill out of a house. They get into a narrow doorway and need to disassemble it before they can continue. Should they reassemble the treadmill as soon as they get through the doorway? Why or why not? What does this have to do with IP datagrams?

  7. A datagram has a length of 4500 bytes and it reaches a link with a MTU of 1500 bytes. How many fragments will it be broken into? (Careful.)

  8. Give the length, fragflag, and offset for each of the datagrams resulting from the scenario in the previous question.

  9. An IP receiver receives packets with the following fields:

    length |  ID | fragflag | offset |
   --- | --- |      --- |    --- |
  1500 |   0 |        1 |      0 |
  1500 |   1 |        0 |    185 |

    Are there any fragments that have not yet been received? If so, what values for these fields would the missing packet(s) need to have to be sure everything was received? If not, how can you tell?

  10. Imagine you are in elementary school and want to know if some like-likes you. If the person you want to ask is in your classroom, you can pass them a note yourself. Otherwise, you need to hand it to the hallway monitor and hope they pass it on for you. What is the networking analogy for this situation?

  11. Even though it is just one device, a router will generally have more IP addresses than an end host. Why?

  12. Consider the host 223.1.4.8 on a subnet identified by 223.1.0.0/16. Which part of the IP address specifies the subnet and which part specifies the host on that subnet? How many hosts can there be on that subnet?

  13. Host A is on a subnet with subnet mask ending in /24. Host B is on a subnet with subnet mask ending in /8. Which subnet has the potential to be bigger (i.e., has more available addresses for hosts on that subnet)?

  14. Why are IPv4 addresses represented as combinations of numbers from 0 - 255? I.e., why is 320.2.900.4 not a valid IP address? Is there any other way that IP addresses could be written besides [0-255].[0-255].[0-255].[0-255]?

  15. Host A has IP address 200.90.1.5. Host B has IP address 200.90.2.5. Can Host A contact Host B without going through a router? Why or why not?

  16. Consider two end hosts connected by a chain of five routers. If you were to sketch out this situation, what is the minimum number of subnets that would be included in your drawing?

  17. With the subnet mask 255.255.252.0, what is the maximum number of hosts on the subnet? If the mask were written in the other format, a.b.c.d/x, what would be the value of x?

  18. With the subnet mask above, find two addresses that have the same first two numbers (a and b) but would reside on different subnets. Then, find two addresses that have the same a and b, have different c, and reside on the same subnet as one another.

For the next few questions, consider the following scenario. A network administrator used to need to get every device hooked up to the company wifi manually. Anyone who came on the property would bring their laptop or phone to them, and they would look at their list of available IP addresses and type one in to the device. Being lazy, like all programmers, the admin set up a new system where people would instead send an email with subject line IP Please as soon as they arrived to the office, and the admin's computer was programmed to automatically reply with an available IP address.

  1. What protocol is the admin trying to reinvent? (Or, in other words, what protocol should they be using instead?)

  2. This scenario will not work as intended by the administrator. Why?

  1. Due to COVID, a major fast food chain starts using a take-a-ticket system so that customers do not need to stand near one another in line. As one of their first customers that day, you get ticket #7. When you go back a few months later, you are given ticket #57934. This large number is kind of inconvenient to say, write down, etc. What lesson from networking did the fast food company fail to take in?

  2. Why is it important that DHCP requests and replies include a transaction ID?

  3. Why are DHCP addresses leased as opposed to given permanently?

For the next two questions, consider a lazy host that does not want to send a DHCP discover message. Instead, it listens for a DHCP offer message that is triggered by someone else's discover message.

  1. Why is it possible for the lazy host to do this in the first place (i.e., why is it able to learn what address was offered to another host)?

  2. If the lazy host sends a DHCP request for that offered IP address, and the original host also sends a request for that address, are there now two hosts that both believe they have the same IP address? If so, what problems does this cause? If not, why not?

  3. ISP A advertises that it services hosts 180.23.0.0/16. ISP B advertises that it services hosts 180.23.17.0/24. Where will packets to the following destinations be routed?

    • 180.23.16.5
    • 180.23.17.2

  4. An ISP owns 8192 IPv4 addresses. Which of the following will make for simpler routing tables elsewhere on the internet?

    • The ISP owns one chunk of 8192 addresses
    • The ISP owns four chunks of 2048 addresses

    Why?

  1. Consider a subnet 10.0.0.0/24 residing behind a NAT. The internet-facing IP address of the gateway router is 270.33.9.4. Host 10.0.0.2 sends a DNS query (port 53) to 138.75.90.2. Give the source IP, source port, destination IP, and destination port of the resulting IP datagrams at each of the following points. If we do not have enough information to determine one of these numbers, choose a random number that is sensible.

    • DNS query while within subnet
    • DNS query while in public internet
    • DNS response while in public internet
    • DNS response within subnet

  2. Why is it possible (sort of) for two hosts on the internet to have the IP address 192.168.2.2?

  3. What is the main reason for the transition to IPv6? Why might routers be able to handle IPv6 packets more efficiently than IPv4 packets?