Quiz 6 - L3 (fragmentation + routing)

Question 1

IPv4 Header fields:
- version
- Internet Header Length (IHL)
- Type of service (ToS)
- Total Length
- Identification
- Flags
- Fragment offset
- Time to Live
- Protocol
- Header Checksum
- Source IP address
- Destinationn IP Address
- Options

Answer 1

• version: format of IP header (4 for IPv4)
• Internet Header Length (IHL): length of header in 32-bit words
• Type of service (ToS): use for DiffServ and ECN
• Total Length: entire packet size in bytes
• Identification: identify fragments of an OG IP datagram
• Flags: 0 - Reserved; 1 - Don’t fragment; 2 - More fragment
• Fragment offset: indicated where in OG datagram this fragment belongs (in 8-bytes units)
• Time to Live: Lifetime of packet. Each router decrements. When it reaches 0, packet is discarded
• Protocol: next protocol used in data portion of IP datagram
• Header Checksum: error detection. recalculated at each router when TTL changes
• Source IP address
• Destination IP Address
• Options

Question 2

For a datagram that is 4000 bytes (including 20-byte header), what is the max data payload when the MTU is 1500 bytes?

Answer 2

Fragment 1:
- length: 1500 bytes
- data: 1480 bytes
- offset: 0
- MF: 1

Fragment 2:
- length: 1500 bytes
- data: 1480 bytes
- offset: 1480
- MF: 1

Fragment 3:
- length: 1020 bytes
- data 1000 bytes
- offset: 1480 + 1000
- MF: 0

Question 3

What is IP fragmentation?

Answer 3

Breaking large IP datagram into small pieces to fit links w/ small MTUs

Question 4

What are the fragmentation fields in the IP header?

Answer 4

• Identification: identify fragments of an OG IP datagram
• Flags: 0 - Reserved; 1 - Don’t fragment; 2 - More fragment
• Fragment offset: indicated where in OG datagram this fragment belongs (in 8-bytes units)

Question 5

Why is fragmentation inefficient?

Answer 5

• header increases overhead
• reassembling fragments is complex (buffering + handling timeouts)
• routers and endpoints have to do more work
• lost fragments cause entire datagram to be retransmitted

Question 6

How does path MTU discovery work to find optimal packet sizes?

Answer 6

The sender sets DF; if a router can’t forward due to MTU, it returns an ICMP “Fragmentation Needed” message.

Question 7

What is ICMP?

Answer 7

Internet control message protocol
- supporting protocol that provides error reporting and diagnostic capabilities for IP.

Question 8

What is the protocol number for ICMP in IP?

Answer 8

1

Question 9

How does ICMP provide encapsulation?

Answer 9

ICMP messages are carried inside IP packets

Question 10

What is a limitation of ICMP?

Answer 10

unreliable (no ACKS or retramsmission)

Question 11

What does the ICMP application ping do?

Answer 11

• tests connectivity and measures RTT
• reveal indo about network path
• uses Echo request/reply

Question 12

How does traceroute use ICMP?

Answer 12

• Sends packets with increasing TTLs; routers reply with ICMP Time Exceeded when TTL = 0.
• reveals path packets take through network

Question 13

How does Path MTU discovery use ICMP?

Answer 13

• Destination Unreachable, Code 4 (“Fragmentation Needed”).
• send packets w/ DF set
• routers that need to fragment send ICMP error w/ next-hop MTU
• can determine largest packet that can travel through the link w/o fragmentation

Question 14

What is the four-setup DHCP process?

Answer 14

Discover → Offer → Request → Acknowledge.

Question 15

What is the purpose of the ARP protocol?

Answer 15

Map IP addresses to MAC addresses on a local network

Question 16

What is the purpose of Dynamic Host Configuration Protocol (DHCP)?

Answer 16

• automatically assign IP addresses and network configuration

Question 17

What problem does NAT solve?

Answer 17

• not enough public IPv4 addresses for all devides
• allows multiple devices to share single public IP address
• Security

Question 18

How does NAT work?

Answer 18

Changes source IP and port for outgoing packets and destination IP and port for incoming packets.

Question 19

Why does NAT require checksum recalculation?

Answer 19

Because NAT changes header fields, affecting checksum values.

Question 20

What are limitations of NAT?

Answer 20

• difficult to initiate connections to internal hosts
• more processing overhead at router that does NAT

Question 21

What is an autonomous system?

Answer 21

• collection of netwroks and routers under the control of a single org.

Question 22

What is the difference between routing and forwarding?

Answer 22

Forwarding is moving packets; routing is deciding paths and building forwarding tables.

Question 23

What is a forwarding table?

Answer 23

set of pairs containing network prefix and next-hop IP address (direct or next-hop)

Question 24

What is Longest Prefix Match?

Answer 24

A rule for forwarding where the route with the longest matching prefix is chosen.

Question 25

How do hosts and routers handle incoming packets differently in IP forwarding?

Answer 25

Hosts: - Check if the packet’s destination address matches their own. - Accept the packet if it matches; otherwise, discard it. Routers: - Look up the packet’s destination address in the forwarding table using longest prefix match. - If a match is found, forward to the specified next hop or interface. - If no match is found, forward using a default route (if available) or discard the packet.

Question 26

How does Distance-Vector Routing work?

Answer 26

Routers share vectors of costs to all destinations with neighbors (Bellman-Ford algorithm).

Question 27

What is convergence?

Answer 27

process of getting consistent routing info to all nodes

Question 28

how is convergence handled with distance-vector protocols?

Answer 28

- update periodically (30 secs) to indicate node is still alive - triggered update - send immediately when topology changes

Question 29

What is the count-to-infinity problem?

Answer 29

A routing loop issue where incorrect paths increase metric values indefinitely.

Question 30

How is count-to-infinity mitigated?

Answer 30

- Split horizon (don't advertise routes back to neighbor from which they learned) - small infinity (e.g., setting 16 as unreachable in RIP).

Question 31

What is link rate routing?

Answer 31

if every node knows the state of evert link, then every node can compute optimal routes

Question 32

What is reliable flooding of link-state info?

Answer 32

process of distributing Link-State Packets (LSPs) to all routers. - Each router sends LSPs to neighbors. - Neighbors forward them to others (except the sender). - Sequence numbers prevent duplicates; TTL limits spread.

Question 33

What’s inside a Link-State Packet (LSP)?

Answer 33

Node ID, neighbor list with link costs, sequence number, and TTL.

Question 34

How do Distance-Vector and Link-State routing compare?

Answer 34

Information shared: - Distance-Vector: Distance to all nodes - Link-State: State of directly connected links Shared with: - Distance-Vector: Immediate neighbors - Link-State: All routers in the network Convergence speed: - Distance-Vector: Slower - Link-State: Faster Memory and computation: - Distance-Vector: Lower resource use - Link-State: Higher resource use Loop prevention: - Distance-Vector: Split horizon, small infinity - Link-State: Full network view prevents loops Scalability: - Distance-Vector: Limited - Link-State: More scalable (supports hierarchy)

Question 35

Name five routing metrics.

Answer 35

Hop count, bandwidth, delay, load, reliability.