Quizzes

Which of the following statements are true about TCP? (all that apply) - TCP offers in order delivery of the packets, flow control and congestion control. - TCP detects packet loss using timeouts and duplicate acknowledgements.

- TCP offers in order delivery of the packets, flow control and congestion control. - TCP detects packet loss using timeouts and duplicate acknowledgements.

(T/F) Consider the TCP Reno, congestion window is cut in half in the event when we either have a timeout, or when we have a duplicate acknowledgement.

False

Assume that router R-D2 learns about a route to a destination in AS-A. How would router R-D2 disseminate this route to R-D3 and R-D1? (Image Needed) Using iBGP. Using eBGP. Using IGP.

Using iBGP.

Which is of the following statements are true about the UDP protocol? (all that apply) - UDP is considered more lightweight than TCP. - When two hosts use UDP to send and receive messages, they need to signal the end of sending data to each other when they are done. - One of the functionalities that UDP offers is to increase or decrease the pace with which the sender sends data to the receiver. - UDP offers basic error checking. -Assume Hosts A, B, C. Host A has a UDP socket with port 123. Hosts B and C each send their own UDP segment to Host A. Hosts B and C cannot use the same destination port 123 for sending their UDP segment.

- UDP is considered more lightweight than TCP. - UDP offers basic error checking.

Consider a router with the forwarding table: Prefix Match Output Link 11100000 00* A 11100000 01000000* B 1110000* C 11100001 1* D otherwise E Given that the router uses longest prefix matching, determine the output link for packet with destination IP address: 11100001 10000000 00010001 01110111

D

Consider a TCP connection, and a diagram that shows how the congestion window progresses over time. From the diagram we can identify the time periods of slow start when the congestion window increases by 1 every RTT.

False

(T/F) Hosts, routers, switches and middlebox devices are expected to implement all layers of the Internet protocol stack.

False Routers are running on network layer whereas the end devices like laptop or iPhones

(T/F) Since the BGP path selection process is fixed, an AS has no control over which routes are selected.

False, AS has control on attributes to chose which routes are selected.

(T/F) An Autonomous System operates across multiple administrative domains.

False, by definition AS is being run by a single AD but a single AD like a CDN can have multiple AS's

(T/F) The Internet topology has been evolving to an increasingly prominent hierarchical structure.

False, going towards flat due to IXP's and CDN's prevalence

(T/F) In TCP, the number of unacknowledged segments that a sender can have is the minimum of the congestion window and the receive window.

True

(T/F) The goal of the spanning tree algorithm is to have bridges select which ports to use for forwarding in order to eliminate loops.

True

Consider a TCP connection, and a diagram that shows how the congestion window progresses over time. From the diagram we can identify the events of packet loss when the congestion window drops to the initial value (usually 1 packet).

True

(T/F) Flow control is a rate control mechanism to protect the receiver's buffer from overflowing. (T/F) Congestion control is a rate control mechanism to protect the network from congestion.

True True

(T/F) The Routing Information Protocol (RIP) is based on the Distance Vector protocol.

True, OSTP is Link-state

(T/F) The forwarding functions of a traditional router refer to transferring packets from the input ports to the appropriate output ports.

True, router is looking at dest of ip and determining appropriate output port. Consults forwarding table

(T/F) The control plane functions can either be implemented in the router's processor or they can be "outsourced" for implementation at a remote controller.

True, this is where software-defined paradigm comes in, implement by outsourcing control plane functions to different device

(T/F) According to the hot potato routing technique, it is at the network's best interest to route the traffic so that it exits the network as soon as possible.

True, we want traffic to be exiting as early as possible. Don't want routers busy and consuming a lot of resources

Which if the following statements are true about the transport layer? (all that apply) - The UDP and TCP protocols have a large overlap of functionalities. - The transport layer protocols offer a logical connection between hosts given that the hosts reside in the same network. - A sender host receives a message from the application layer it encapsulates it with the transport layer header before passing it down to the network layer. - An application running on a host can bind to multiple sockets simultaneously. - A host can't maintain a TCP socket and a UDP socket simultaneously.

- A sender host receives a message from the application layer it encapsulates it with the transport layer header before passing it down to the network layer. - An application running on a host can bind to multiple sockets simultaneously. TCP has flow control and in order delivery, they do not have many overlapping functionalities.

In traditional routers, traffic forwarding is performed based on: - Destination IP address only - Source IP address only - Both source and destination IP addresses

- Destination IP address only

Consider the figure below: Note: C1, C2, C3 are customers of ISP-X. ISP-P is a provider of ISP-X. (Image Needed) Assume that AS-X learns multiple routes for the same external destination W. These multiple routes are advertised from C3, from Y and from P. How would AS-X rank these routes before deciding which one to import? A. 1) route learned from C3, 2) route learned from Y, 3) route learned from P. B. 1) route learned from Y, 2) route learned from C3, 3) route learned from P. C. 1) route learned from P, 2) route learned from Y, 3) route learned from C3.

A. 1) route learned from C3, 2) route learned from Y, 3) route learned from P. We know that it will prefer routes customer (provides revenue), peers (free traffic), providers (last cause will result in additional costs)

Consider a router with the forwarding table: Prefix Match Output Link 11100000 00* A 11100000 01000000* B 1110000* C 11100001 1* D otherwise E Given that the router uses longest prefix matching, determine the output link for packet with destination IP address: 11100001 01000000 11000011 00111100

C

Consider a router with the forwarding table: Prefix Match Output Link 11100000 00* A 11100000 01000000* B 1110000* C 11100001 1* D otherwise E Given that the router uses longest prefix matching, determine the output link for packet with destination IP address: 10001000 11110001 01010001 11110101

E no prefix match starting

(T/F) At an IXP, the members have the choice to peer privately or publicly.

True, this is a service option

Assume that router R-D2 learns about a route to an internal destination in AS-D. How would router R-D2 disseminate this route to R-D3 and R-D1? (Image Needed) Using iBGP. Using eBGP. Using IGP.

Using IGP. Route traffic within an AS then it would not include BGP. Internal destinations uses IGP

Consider the following unibit trie. What is the longest prefix match for 111*?

add up [01100000] B2

Consider the following unibit trie. What is the longest prefix match for 11011*?

add up [1011010000] P9, refer to last successful cause 0 points to it

(T/F) Hubs can be used interchangeably with layer-2 switches.

False

(T/F) For two ASes to form a peering agreement, they need to find common ground regarding the internal policies and traffic engineering approaches that each AS implements.

False, when having 2 AS's with peering agreement it is a business agreement. Roughly the same amount of traffic where we don't need to pay each other but in addition may have additional terms (think SLA's). Other than that one AS does not dictate what happens internally in another AS

What are the services that TCP provides? (all that apply) - Flow control - Forwarding - Congestion control - In-order delivery of packets - Routing

Flow control Congestion control In-order delivery of packets TCP is running in the transport layer. Forward and routing are not in the transport layer they are in the network layer

Consider the topology below and the types of BGP sessions the routers form. (Image Needed) Select the true statements. R-A1 <-> R-A3 : iBGP R-B1 <-> R-A3 : eBGP R-B3 <-> R-D2 : iBGP R-D3 <-> R-D2 : eBGP

R-A1 <-> R-A3 : iBGP R-B1 <-> R-A3 : eBGP Within AS - iBGP Between AS - eBGP

Which if the following statements is true about learning bridges? - A network operator needs to configure the forwarding table. - A learning bridge populates the forwarding table based on the network traffic conditions such as volume and traffic disruptions. - The forwarding table represents which hosts are reachable through which ports.

The forwarding table represents which hosts are reachable through which ports.

(T/F) By stride we refer to the number of bits that we check at every step when traversing a trie.

True

(T/F) The identifier of a UDP socket is a tuple of destination IP address and port. (T/F) The identifier of a TCP socket is a tuple of source IP address and port.

True False

Which if the following statements are true? - Routing and forwarding are interchangeable terms. - Assume a source and a destination host. Before packets leave the source host, the host needs to define the path over which the packets will travel to reach the destination host. - Intradomain routing refers to routing that takes place among routers that belong to the same administrative domain. In contrast when routers belong to different administrative domains we refer to routing as interdomain routing.

- Intradomain routing refers to routing that takes place among routers that belong to the same administrative domain. In contrast when routers belong to different administrative domains we refer to routing as interdomain routing. routing and forwarding are not same terms, source hosts don't know the path - not determined by the end points

Consider the distance vector (DV) routing algorithm. Which of the following statements are true? (all that apply) - It keeps iterating as long as neighbors send new updates to each other. - It is centralized. - It requires synchronization between routers. - Each node maintains and updates its own view of the network. - The distance vector, that each node maintains, is essentially a table with costs to reach every node in the network. - Each node uses the Bellman Ford equation to update its distance vector.

- It keeps iterating as long as neighbors send new updates to each other. - Each node maintains and updates its own view of the network. - The distance vector, that each node maintains, is essentially a table with costs to reach every node in the network. - Each node uses the Bellman Ford equation to update its distance vector. it is decentralized, distributed algorithm, and asynchronous

Which of the following statements are true about TCP Cubic? - TCP Cubic was designed for better network utilization, and lower network congestion. - TCP Cubic congestion window growth function is designed to not overflow the receiver's buffer. - TCP Cubic uses a cubic function to increase the congestion window. - TCP Cubic increases the congestion window in every RTT.

- TCP Cubic was designed for better network utilization, and lower network congestion. - TCP Cubic uses a cubic function to increase the congestion window.

Consider the link state routing protocol. Which of the following statements are true? (all that apply) - The link costs are known to all nodes. - The network topology is known to all nodes. - Consider a node u as our source node, the goal of the algorithm is to compute the least cost paths, from the source node u to every other node in the network. - Consider a node u as our source node. Consider the initialization step of the algorithm. We initialize the least cost path to directly attached neighbors to be the cost of the direct links, and for non-directly attached neighbors we initialize the least cost path to infinity.

- The link costs are known to all nodes. - The network topology is known to all nodes. - Consider a node u as our source node, the goal of the algorithm is to compute the least cost paths, from the source node u to every other node in the network. - Consider a node u as our source node. Consider the initialization step of the algorithm. We initialize the least cost path to directly attached neighbors to be the cost of the direct links, and for non-directly attached neighbors we initialize the least cost path to infinity.

Select which of the following statements are true about the Evolutionary Architecture Model (or EvoArch): (all that apply) - The model suggests that the IP protocol survived because the protocols at the transport layer have been stable themselves. - The TCP/IP stack did not compete with pre-existing network services, instead it targeted new applications. - The model predicts that even if new Internet architectures are not initially designed to have the shape of an hourglass, they will probably do so as they evolve over time.

- The model suggests that the IP protocol survived because the protocols at the transport layer have been stable themselves. - The TCP/IP stack did not compete with pre-existing network services, instead it targeted new applications. - The model predicts that even if new Internet architectures are not initially designed to have the shape of an hourglass, they will probably do so as they evolve over time.

(T/F) The evolution of the Internet protocols has shown that most of the innovation has been happening at the protocols/layers that are at the "waist" of the protocols hourglass.

False

(T/F) Let's assume that you design a new application, and you choose to build it on top of UDP. You can't incorporate additional services at the application layer itself, that are not offered by UDP.

False So for example, UDP does not offer reliable transmission. There is nothing saying that you cannot build your application to support reliable transmission, implemented at a higher level of the OSI model. UDP would continue to operate like it does, but you could have checking of some type at a higher level. Make sense? You can implement application however you want A good example of this is TFTP. Regular FTP uses TCP (ports 20/21, if I remember correctly). TFTP uses UDP (port 69) and implements ACKs at the application layer.

(T/F) Let's assume a packet that traverses a path from a source host to the destination host. As the packet traverses the path, each node de-encapsulates the packet first before it sends it over the physical layer towards the next node.

False To clarify *just before it sends it over layer to physical node*. What happens to the packet as it's sent from one node to the next.

(T/F) The DV count to infinity problem states that good news (eg a decrease in a link cost) propagates slow among nodes in the network.

False, DV says that good news propagates fast and bad news propagates slow

(T/F) There are no costs involved for an AS to participate at an IXP.

False, charges for speed, other value added services

(T/F) Every network can only have a specific number of egress points.

False, does not have to have a predetermined number of egress points

(T/F) The poison reverse technique solves the count to infinity problem for all network topologies.

False, does not solve count to infinity problems that include 3+ non-connected nodes

(T/F) Assume a source and a destination host. As packets travel over a path from the source host to the destination host the packets are handled by multiple routers over that path. If these routers belong to different administrative domains they need to run the same intradomain routing algorithm, since they are on the same path for that pair of hosts.

False, internal routing is determined by AS, each is independent

(T/F) In the case that a network has multiple egress points, then the egress points should offer paths to external destinations with different costs.

False, it does not have to be with different costs, it can be but does not have to. Those egress points offere two different points to exit network. Can be associated with different intra-domain paths. There may be other reasons why we want multiple exits

(T/F) Since local traffic stays local at IXPs, the IXP infrastructures deal with limited volumes of traffic.

False, local traffic stays local but that still means networks are communicating. Traffic doesn't need to go through intermediate devices. This doesn't change amount of traffic. Large participants in IXP's can deal with huge amounts of traffic

(T/F) An IXP route server does not need to run the BGP protocol to facilitate the establishment of multi-lateral peering sessions.

False, route server needs to run BGP independently with every participant AS.

(T/F) IXPs leading incentive to establish route servers was to charge the participants for using it.

False, their initial incentive was to facilitate peering of participant networks. Can do it in a bilateral way 1-1 relationship with each network. More efficient to establish multi-lateral peering through IXP

(T/F) For multi-lateral BGP peering sessions at an IXP, the participants have the choice to advertise routes, either directly to other participants, or to the route server.

False, they don't have this choice. This only happens with route server. 1-1 bilateral to do that

(T/F) Consider the figure below that denotes ASes and their relationships. Note: C1, C2, C3 are customers of ISP-X. ISP-P is a provider of ISP-X. (Image Needed) ISP-X has the incentive to advertise routes for C3 to Y.

ISP-X has the incentive to advertise routes for C3 to Y. Image shows the map of the agreements. They have a peering agreement between ISP X and Y, so X tells Y the customers and vice versa [t]

(T/F) Consider the figure below that denotes networks and their relationships. Note: C1, C2, C3 are customers of ISP-X. ISP-P is a provider of ISP-X. (Image Needed) ISP-X has the incentive to advertise the routes for P's customers to Y and Z.

ISP-X has the incentive to advertise the routes for P's customers to Y and Z. If ISP X advertises P's routes to Y and Z this will result in additional costs for ISP X because they have customer provider relationship between X and P so all traffic between them results in additional costs [f]

Which of the following statements is true for the end to end principle? (choose one) - The principle suggests to keep the network core minimal and offload the intelligence to the end systems. - Given that the network functions are changing, the principle suggests to build an intelligent network core so that it can readily support an increasing variety of challenges and applications. - The principle has been blamed to delay the proliferation of applications. - A middlebox device cannot be deployed at a network unless it is designed based on the principle.

The principle suggests to keep the network core minimal and offload the intelligence to the end systems.

(T/F) An AS can use LocalPref to control which routers are used as exit points (for the outgoing traffic), and it can use the MED attribute to control which routers are used as entry points (for the incoming traffic).

True,

(T/F) The forwarding plane operations take place in shorter time scales than the control plane operations.

True,

(T/F) When a large provider or Content Delivery Network joins an IXP, this can act as an incentive for other networks to join as well.

True,

(T/F) A Content Distribution Network (CDN) or an ISP can operate over multiple Autonomous Systems.

True, a AD like CDN on other networks can have multiple AS's

(T/F) The data plane functions of a traditional router are implemented in hardware.

True, actual forwarding happens here

(T/F) Assume that AS-B learns about an external destination both from AS-C and from AS-A. AS-B can show preference to use the route heard from AS-C by assigning higher LocalPref value to that route.

True, by doing this network operator of AS-B is controlling how traffic will enter a network. Same external destination and by assigning it a higher value it lowers the preference

(T/F) Assume that AS-B advertises the routes to its internal destination to AS-A using the routers R-B1 and R-B4. AS-B can communicate to AS-A that it prefers R-B1 as an entry point to the network, by assigning lower MED values to these routes.

True, communicating with neighboring network and can assign a lower MED value meaning it is prefers it

(T/F) The control plane functions of a traditional router are implemented in software.

True, intelligence happens here

(T/F) One of the services provided by IXPs is additional security protections such as mitigation of DDoS (Distributed Denial of Service) attacks.

True, it is an an additional service IXP's are offering

(T/F) In a router, there is no interplay between the control plane and the forwarding plane to compute the forwarding table.

True, only routing protocols take place in control plane

Assume that router R-B1 learns about a route to AS-C. How would router R-B1 disseminate this route to R-A3? (Image Needed) Using iBGP. Using eBGP. Using IGP.

Using eBGP.