The class has decided to write the midterm themselves. These are the approved questions. The midterm will be taken from these questions. Exam questions may have different wording, but the subject of every question on the exam will match a subject listed here.
| Topic | Question |
|---|---|
| UDP | Design a simple protocol using UDP that can handle file transfers across a network, ensuring the file is reconstructed at the destination correctly. |
| TCP | Explain the difference between TCP’s congestion control and a simple UDP-based protocol, considering how you might implement a custom congestion control mechanism over UDP. |
| IPv6 | Investigate how IPv6 solves issues such as address exhaustion compared to IPv4 |
| packet size | Outline the steps for diagnosing a network issue where large packets are not successfully reaching their destination. |
| IXP | What are IXPs and how do they work? |
| TTL | Describe two experiments to determine TTL for a network system |
| AS | What is an autonomous system and how does it work? |
| Principles | What are five sensible design principles when designing a network? |
| smart–dumb network | If the Internet was a smart network rather than a dumb one, how might it operate? |
| ethernet | What influence will it have on Ethernet transmission if there was no arbitrary backoff used by transceivers following collisions? |
| GCP | What role does the Good Citizen Principle play in the management of Internet resources when demand is high? |
| transport layer | How does the transport layer confirm that data is accurately transmitted across the internet? |
| router info exchange | What distinguishes link-state routing from distance-vector routing and makes it more scalable for larger networks? |
| port | Where are IP ports located on a PC? |
| DHCP | how is DHCPv6 different from DHCPv4 in terms of network configuration management |
| ethernet | How does the Ethernet protocol handle collisions? |
| MAC address | How is a MAC address different from an IP address? |
| MAC address | What happens if two devices have the same MAC address on the same network? |
| MAC address | What happens if two devices have the same MAC address on different networks? |
| subnet | What does /24 mean in 192.168.1.1/24? What would the subnet mask for this address space be? |
| If I want to send an email to my aunt in Moscow, describe step by step how a packet from my house could cross the world and get to her computer? | |
| socket | What is the difference between a socket and a port? |
| subnet | How does a router know what subnet to use? |
| subnet | What does /24 mean in 192.168.1.1/24? What would the subnet mask for this address space be? |
| routing | What is the purpose of a forwarding table in a router, and how does it use IP addresses to determine where to send data packets. |
| ICMP | Can ICMP support IPv4 and IPv6, what were the ICMP messages that you came across while working on Wireshark and explain them. |
| TCP | Explain 3-way handshake and give an example for sliding window with window size 5. |
| RIP | Explain RIP for IPv4 and IPv6. |
| router protocol | In a network where some routers participate in a routing protocol and others do not, explain how the ‘extra hop problem’ can occur when nonparticipating routers use participating routers as a default route. |
| BGP | Because BGP uses a reliable transport protocol, KEEPALIVE messages cannot be lost. Does it make sense to specify a keepalive interval as one-third of the hold timer value? Why or why not? |
| UDP | How can an application ensure messages sent over UDP arrive correctly without using TCP? |
| MPLS | How does MPLS speed up data forwarding compared to traditional IP routing |
| RARP | If a diskless workstation can’t use RARP to find its IP address, what problems might it face? |
| ARP | f ARP couldn’t match IP addresses to MAC addresses, how would devices communicate on a local network? |
| multicast | What issues might arise if multicast routing protocols used the same algorithms as unicast routing? |
| MPLS | What would be the consequences if MPLS were used in place of traditional IP routing without using labels? |
| broadcasting | How would network performance be affected if all devices used only broadcasting instead of switching for internal communications? |
| multicast | When one uses IP network, what are the differences understood between unicast, broadcast and multicast. |
| RPF | What network performance problems would arise from missing out on RPF(reverse path forwarding) in multicast networking? |
| MPLS | If MPLS decided to handle unicast and multicast traffic to what extent would changes be made to the label switching and path creation as well as how it would affect routing system efficiency? |
| IGMP | What changes to the IGMP protocol will be needed to make in case multicast groups stop being dynamic (i.e. are constant and does not change over time)? |
| router time | Why is it important to conserve router time? |
| GCP | Give three examples of the Good Citizen Principle in protocols we have studied. |
| smart–dumb network | Give three advantages of smart networks. |
| smart–dumb network | Give three advantages of stupid networks. |
| UDP | How can I write an application to use UDP for reliable transport? |
| Internet | Where is the headquarters for day-to-day internet operations? |
| router time | Give three examples where protocol design decisions have been made to conserve router time. |
| ICMP | If an ICMP message is not deliverable, what ICMP message is sent to the originator? |
| router protocol | Explain split horizon in router protocols. |
| router protocol | Explain poison reverse in router protocols. |
| smart–dumb network | Are end-to-end protocols a characteristic of smart or dumb networks? |
| TCP | Can I use TCP over a satellite? Why? |
| MPLS | What is the primary benefit of using MPLS over traditional IP routing? |
| MPLS | How would a network behave if a router mistakenly assigned the same MPLS label to different paths? |
| multicast | What could happen if routers failed to poll hosts about their membership in multicast groups? |
| Multicast | How would network efficiency be impacted if multicast used flooding instead of Protocol Independent Multicast (PIM) sparse mode for routing? |
| IGMP | How does IGMP manage group membership in a multicast network? |
| DNS | How could DNS functionality change if all name servers used only iterative resolution instead of recursive resolution? |
| DHCP | What would happen if a DHCP server assigned extremely short lease durations for IP addresses on a busy network? |
| NAT | What is the most significant limitation of capability when NAT is employed? |
| ICMP | What would be the consequence if routers used ICMP Source Quench messages excessively to manage network congestion? And what do routers do about it? |
| ICMP | What does an ICMP Source Quench message do? |
| routing | What would happen if routers did not use split horizon in distance vector routing protocols? |
| TCP | How would TCP work differently if it did not use a three-way handshake to start connections? |
| routing | What might happen if link-state routing protocols were used in all networks, even very small ones? |
| routing | Which is better, link-state or distance-vector? Why? |
| routing | If a router ignored TTL values in IP packets, what problems could happen? |
| UDP | What would happen if UDP used error correction like TCP? |
| IP | How would packet delivery change if IP did not break up large packets? |