Cse 675 Midterm 2

Front 1

Connection-Oriented

Back 1

Befre applications can send data they must handshake with eachother and start the transfer

Front 2

Full Duplex Service

Back 2

Both hosts can send data back and forth to each other at the same time

Front 3

Point to Point

Back 3

Connection between single receiver ansender. Only two hosts

Front 4

How does the three way handshake work

Back 4

1.) Client sends syn segment with initial sequence value
2.) Server sends synack with initial sequence value
3.) Client sends ack with possible data

Front 5

Send buffer

Back 5

On the sender side, application data is stored here and sent intermittently

Front 6

MSS

Back 6

Maximum Segment Size, the maximum amount of data in a tcp segment. Not including header information

Front 7

MTU

Back 7

Maximum Transmission Unit - Largest link layer frame that can be sent by sending host

Front 8

TCP Segment

Back 8

Combination of data and header information

Front 9

Sequence numbers in TCP

Back 9

Are based on bytes of data and what the receiver is expecting to get next

Front 10

What is the receive window used for

Back 10

flow control on the receiver side

Front 11

How do ACKS work in TCP

Back 11

Send ack for next byte you are expecting. Cumulative

Front 12

RTT

Back 12

round trip time, time from when a segment is sent until it is acknowledged

Front 13

How is Sample RTT taken?

Back 13

Oldest segment is timed for RTT. Not affter retransmit

Front 14

Timeout Interval equation

Back 14

EstimatdRTT + 4 * DevRTT

Front 15

Fast Retransmit

Back 15

After receiving triple ACK, the sender will resend Acked segment before the timer goes of

Front 16

Selective Acknowledgment

Back 16

Allows a receiver to acknowledge out of order segments selecitvly rather than just cumulatively acknowleding the last correctly received in order segment

Front 17

Receive Window

Back 17

Sender variable that gives an idea to sender about how much free buffer space is available at the receiver

Front 18

Receive Buffer

Back 18

The amount of data a client is willing to receive for this connection

Front 19

Last Byte Read

Back 19

The number of the last byte in the data stream read from the buffer by the client process

Front 20

LastByteRcvd

Back 20

The number o the last byte in the data stream that has arrived from the network and has been placed in the receive buffer at the client

Front 21

LastByteSent

Back 21

The last byte the sender has sent

Front 22

LastyteAcked

Back 22

The last byte the sender has received an ack for

Front 23

What does the server make sure of for flow control

Back 23

LastByteSent - LastByteAcked < rcvwindow

Front 24

What does the client do when its rwnd becomes 0 and it has nothing to send to the server?

Back 24

Server sends 1 byte data (with seq of ack) packages to client so client acks with rcvwindow size

Front 25

How is a TCP connection initiated

Back 25

1.) Client sends SYN packet with no data and segment it wants to start at
2.) Server allocates variables. Respons with SYN, ACKS segment #+1 and sends own segment
3.) Client allocates variables sends regular data with ack set to client + 1

Front 26

How is a TCP connection closed

Back 26

1.) Client sends Fin
2.) Server Acks
3.)Server sends Fin
4.) Client Acks
5.) Client waits for time after time is up, clears connection

Front 27

Offered load

Back 27

The rate at which the transport layer sends segments (containing original and retransmitted data) into the network

Front 28

End-to-End congestion Control

Back 28

Networ layer provides no explicit support and the end system must control it

Front 29

Network-Assisted Congestion Control

Back 29

Routers provide feedback to the sender regarding congestion.

Front 30

What are the two ways Network-assisted congestion control work

Back 30

1.) Choke Packet - Router sends information to process
2.) Router modifies a bit in packet

Front 31

congestion window

Back 31

A sender side value that constrains the rate at which a sender can send traffic into the network

Front 32

slow start

Back 32

cwnd starts at 1 and increases by 1 for each received ack

Front 33

3 Ways to leave slow start

Back 33

1.) CWND >= ssthresh
congestion avoidance

2.) Timeout. Start slow start but set cwnd to 1 and ssthresh to cwnd/2

3.) 3 duplicate acks. Set ssthres to cwnd/2 and cwnd to ssthresh + 3 and go to faast recovery

Front 34

When is congestion avoidance entered?

Back 34

1.) When CWND >= ssthresh

2.) New ack is received when in fast recovery

Front 35

How does congestion avoidance work

Back 35

For each new ack, cwnd is incremented by cwnd + mss * (mss /cwnd)

Front 36

What happens for congestion control when 3 duplicate acks are encountered

Back 36

ssthresh = cwnd /2
cwnd = ssthresh + 3

Front 37

What happens in fast recovery / how do get out

Back 37

1.) Every duplicate ack increases cwnd by mss

EXITING

2.) ON timeout go to slow start
ssthresh = cwnd/2
cwnd = 1

3.) ON new ack go to congestion
cwnd = ssthresh

Front 38

What is the difference between tahoe and reno

Back 38

TCP reno has fast recovery after a triple ack, tahoe goes to slow start stage

Front 39

AIMD

Back 39

Additive-increase, multiplicative-decrease. Sawtooth graph

Front 40

Forwarding

Back 40

Local action of transferring a packet from an input link interface to the appropriate output link interace

Front 41

Routing

Back 41

Network-wide process that determines the end-to-end paths that packets take from source to destination

Front 42

Forwarding Table

Back 42

Router reads the value in the packets header and determines which out going link to put the packet to

Front 43

How are forwarding tables populated (3 things)

Back 43

1.) By routing algorithms

2.) Central cerver routing alogortihms

decentralized routing algorithms

Front 44

Link Layer Switches

Back 44

Base forwarding decision on value in link-layer field

Front 45

Routers (forwarding decisions)

Back 45

Base decision on network layer-field

Front 46

Network Service model

Back 46

The characteristics of end-to-end transport of packets between sending and receiving end systems

Front 47

best effort service

Back 47

1.) Timing between packets isn't guranteed

2.) Packet order isn't guranteed

3.) Delivery of packets isn't guranteed

Front 48

CBR ATM

Back 48

(Constant Bit Rate) Network service that provides constant bit rate

Front 49

ABR

Back 49

(Avaiable Bit Rate) Minimum throughput provided

Front 50

VC Network

Back 50

Virtual Circuit Networks - provide a connection service at the network layer

Front 51

Datagram Network

Back 51

Provides a connectionless service

Front 52

Signaling messages

Back 52

Messages that the end systems send into the network to initiate or terminate a VC, and the messages passed between the routers to set up the VC

Front 53

Longest Prefix matching rule

Back 53

longest matching entry in the table is the match

Front 54

Switching Fabric ways

Back 54

memory, crossbarr, bus

Front 55

Packet Scheduler

Back 55

Output port must queue requests and determine which to transmit

Front 56

Drop Tail

Back 56

Drop the remove one of already queued packets to make room for a new one

Front 57

HOL

Back 57

Head of line blocking - queued packet in input queue must wait for transfer through the fabric (even though its output port is free) because it is blocked by another packet at the head of the line

Front 58

Why is the header length field needed in the IP datagram

Back 58

Can contain a variable number of options

Front 59

What is the datagram length field in the ip datagram

Back 59

The total length of the datagram (header plus data) in bytes

Front 60

Interface

Back 60

Boundary between the host and physical link

Front 61

Subnet

Back 61

Network that interconectes other host interfaces and one router interface

Front 62

Subnet mask

Back 62

Denoted as ip with /* where all front bits are required

Front 63

CIDR

Back 63

Classless interdomain Routing . Generalized the notion of subnet addressing.

Front 64

Broadcast address

Back 64

255.255.255.255 Broadcasts information to all hosts on a subnet

Front 65

DHCP

Back 65

Dynamic Host Configuration Protocol. Allows a host to obtain an ip address automatically

Front 66

NAT

Back 66

network address translation

Front 67

realm

Back 67

A network whose addresses only have meaning to devices within that network

Front 68

ICMP

Back 68

Internet Control message Protocol

Front 69

TCP Header - Header length Field

Back 69

Since the options field makes it so TCP headers can be variable size, the header field is required to indicate how many bytes are in the header

Front 70

TCP Header - ACK bit

Back 70

Indicates that the value carried in the ack field is valid

Front 71

TCP Header - RST, SYN, FIN bits

Back 71

Used for connetion setup and teardown

Front 72

TCP Header - PSH bit

Back 72

indicates that the receiver should pass the data to the upper layer immediately

Front 73

TCP Header - URG bit

Back 73

used to indicate that there is data in this segment that the sending-side upper-layer has marked urgent

Front 74

RDT 1.0

Back 74

Reliable data transfer over perfectly reliable channel

Front 75

RDT 2.0

Back 75

Bit errors, send ack or nak to determine if we got a corrupted/uncorrupted piece of data

Front 76

Stop and Wait protocol

Back 76

Sender will not send new packet until it is sure previous packet has been received correctly

Front 77

RDT 2.1

Back 77

Add sequence numbers

Front 78

rdt 3.0

Back 78

Can lose packets. add timers

Front 79

Memorize TCP header

Back 79

SDSA
HUFR
IUOD

source port
dest port
sequence number
ack number
header length
unusued
flags
receive window
internet checksum
urgent data point
options
data

Front 80

IPv4 Datagram header - version number

Back 80

4 bits that indicate what version is being used

Front 81

IPv4 Datagram header - header length

Back 81

the length of the variable length header. Indicates where the data begins

Front 82

IPv4 Datagram header - type of service

Back 82

used to distinguish if application is real time or what

Front 83

IPv4 Datagram header - datagram length

Back 83

total length of the datagram (header plus data) measured in bytes
16 bits

Front 84

IPv4 Datagram header - identifier, flags, fragmentation offset

Back 84

used for fragmentation

Front 85

IPv4 Datagram header - ttl

Back 85

time to live, used to ensure datagrams are dropped if this value reaches 0

Front 86

IPv4 Datagram header - protocol

Back 86

used when the datagram reaches its final destination. 17 for udp and 6 for tcp

Front 87

IPv4 Datagram header - header checksum

Back 87

checksum used for the ip datagram header

Front 88

IPv4 Datagram header - Ip addresses

Back 88

indicate where to deliver the datagram

Front 89

IPv4 Datagram header - options

Back 89

allow the header options to be extended

Front 90

IPv4 Datagram header - data

Back 90

where the actual data is held

Front 91

how does CIDR work

Back 91

First portion of bits indicate network. next part indicates inside an organization

Front 92

How are ICMP packets carried

Back 92

As IP data payload

Front 93

How is ICMP read

Back 93

it is demultiplexed to ICMP

Front 94

ipv4 datagram format

Back 94

VHTD
IFF
TUH
S
D
O
D

1.) version, haeder length, type of service, datagram length
2.) identifier, flags, fragmentation offset
3.) time to live, upper layer protocol, checksum
4.) soruce ip
5.) destination ip
6.) options
7.) data