Lecture 6: 5th November 2019

Question 1

What is QUIC?

Answer 1

Quick UDP Internet Connections = a new transport layer Internet protocol to improve upon and replace TCP. It’s faster and encrypted-by-default.

Question 2

How can you increase the speed of the Internet with respect to the application, session, and transport layers of the reference model? How else can this be defined?

Answer 2

!

Question 3

How can you increase the speed of the Internet with respect to the network infrastructure of service providers?

Answer 3

!

Question 4

Why/how is QUIC influential?

Answer 4

!

Question 5

What layers in the reference model does QUIC sit in?

Answer 5

!

Question 6

What does QUIC interface with?

Answer 6

!

Question 7

Are the features of QUIC new? Why is it useful?

Answer 7

!

Question 8

What are the key reasons why we should replace TCP?

Answer 8

!

Question 9

What does it mean to say the application context of HTTP has changed?

Answer 9

!

Question 10

If HTTP/2 fixes application-layer HoL blocking found in HTTP/1.1, then why does it still happen with HTTP/2 flows?

Answer 10

!

Question 11

How does QUIC solve the HoL blocking of HTTP/2?

Answer 11

!

Question 12

How does TCP’s sliding window halt flows?

Answer 12

!

Question 13

What are the issues with TCP’s handshake process?

Answer 13

!

Question 14

How many RTTs are required for a (unencrypted) TCP handshake?

Answer 14

!

Question 15

How many RTTs are required for a handshake for a TCP flow encrypted with TLS1.2?

Answer 15

!

Question 16

How does TLS1.3 improve on the time taken to perform a handshake? How many RTTs does it take to set up a TCP flow with TLS1.3?

Answer 16

!

Question 17

How do initial QUIC handshakes work? How many RTTs doe they take to complete?

Answer 17

!

Question 18

How do subsquent QUIC handshakes work? How many RTTs do they take? What are they officially called?

Answer 18

!

Question 19

Does QUIC affect congestion control?

Answer 19

No

Question 20

What effect does a high latency have on handshake times? How does the number of RTTs taken affect this?

Answer 20

!

Question 21

Why can’t you combine the handshakes of TCP and TLS into 1?

Answer 21

!

Question 22

How do initial handshakes work with TCP and TLS 1.2? How many RTTs do they take?

Answer 22

!

Question 23

How do subsequent handshakes work with TCP and TLS 1.2? How many RTTs do they take?

Answer 23

!

Question 24

How do initial handshakes work with TCP and TLS 1.3? How many RTTs do they take?

Answer 24

!

Question 25

How do subsequent handshakes work with TCP and TLS 1.3? How many RTTs do they take?

Answer 25

!

Question 26

What is a REJ packet?

Answer 26

!

Question 27

What is a REJ packet composed of?

Answer 27

!

Question 28

What is an inchotate CHLO packet?

Answer 28

!

Question 29

What is an inchotate CHLO packet composed of?

Answer 29

!

Question 30

What is a complete CHLO packet?

Answer 30

!

Question 31

What is a complete CHLO packet composed of?

Answer 31

!

Question 32

What is the security flaw of QUIC's handshakes?

Answer 32

replay attacks

Question 33

Why are replay attacks possible due to QUIC's handshakes?

Answer 33

!

Question 34

What are the 3 solutions to preventing replay attacks with QUIC's handshakes?

Answer 34

!

Question 35

Why can you not maintain a global and temporal consistency of QUIC connections on all servers?

Answer 35

!

Question 36

What are idempotent operations?

Answer 36

!

Question 37

Why can you not limit hosts to using idempotent operations with 0-RTT QUIC handshakes to prevent replay attacks?

Answer 37

!

Question 38

How is TCP ill-equipped for mobility?

Answer 38

!

Question 39

How does QUIC handle host IP address changes?

Answer 39

!

Question 40

How does QUIC implement mobility?

Answer 40

!

Question 41

Why is it difficult to retroactively change TCP?

Answer 41

!

Question 42

Why do middleboxes impede Internet functionality?

Answer 42

!

Question 43

How do middleboxes violate the end-to-end principle?

Answer 43

!

Question 44

Why are packets with option headers often discarded by middleboxes?

Answer 44

!

Question 45

How do packet headers change from TCP to QUIC?

Answer 45

!

Question 46

How do the packet changes from TCP to QUIC affect middleboxes?

Answer 46

!

Question 47

What are the pros of QUIC compared to TCP?

Answer 47

!

Question 48

What are the cons of QUIC compared to TCP?

Answer 48

!

Question 49

What is BBR?

Answer 49

Bottleneck Bandwidth and Round-trip propagation time (BBR) is a TCP congestion control algorithm developed at Google in 2016 that uses the maximum bandwidth and round-trip time of the last window to build an explicit model of the network. Higher capacity NICs mean latency/model-based congestion control algorithms, such as BBR, provide higher throughput and lower latency as a more reliable alternative to more loss-based algorithms like CUBIC.

Question 50

What are the basics of how BBR works?

Answer 50

!

Question 51

What are the issues of TCP Reno? What highlighted these problems?

Answer 51

!

Question 52

When are links lossy?

Answer 52

!

Question 53

What is CUBIC?

Answer 53

!

Question 54

What is the purpose of CUBIC?

Answer 54

!

Question 55

How does CUBIC work?

Answer 55

!

Question 56

How are CUBIC's queue size and capacity related?

Answer 56

!

Question 57

What are the states of flow management within loss-based congestion control?

Answer 57

!

Question 58

What are the problems with loss-based congestion control?

Answer 58

!

Question 59

What does BBR use indtead of congestion avoidance?

Answer 59

!

Question 60

How does RTT change as the number of bytes in flight at once increases?

Answer 60

!

Question 61

How does throughput change as the number of bytes in flight at once increases?

Answer 61

!

Question 62

What is the bottleneck bandwidth?

Answer 62

!

Question 63

What is BDP?

Answer 63

!

Question 64

What is the gradient by which throughput increases with the number of bytes in flight at once up to the BDP?

Answer 64

!

Question 65

What is the gradient by which RTT increases with the number of bytes in flight after the BDP?

Answer 65

!

Question 66

What is Kleinrock's optimal operating point?

Answer 66

!

Question 67

How is the optimum throughput found in loss-based congestion control different to Kleinrock's optimal operating point?

Answer 67

!

Question 68

How does the difference between the optimum throughput found in loss-based congestion control and Kleinrock's optimal operating point highlight an issue with loss-based congestion control?

Answer 68

!

Question 69

What are the two competing targets of BBR?

Answer 69

!

Question 70

What is the problem with the two competing targets of BBR?

Answer 70

!

Question 71

Why does BBR try to detect queueing?

Answer 71

!

Question 72

How does BBR try to detect queueing?

Answer 72

!

Question 73

What is the BBR algorithm composed of (give a conceptual overview)?

Answer 73

!

Question 74

How does the sending rate of BBR change over time? How does this compare to Reno and CUBIC?

Answer 74

!

Question 75

How does BBR's sending rate and queue size change over time?

Answer 75

!

Question 76

How are the sending rate and queue size of BBR related?

Answer 76

!

Question 77

What are the key pros of BBR?

Answer 77

!

Question 78

What are the key cons of BBR?

Answer 78

!