Networking Exam 3 Chapter 8
How long is each part of an IP address?
16, 8, 8 bits
How many bits are there in a mask?
32 bits
Know how to express: A mask has eight 1s followed by 0s. Express the mask in DDN. Express the mask in PN.
32-8=24 bits are 0s. Binary form is 11111111 00000000 00000000 00000000. First eight 1s from decimal 255, next eight 0s form decimal 0, next eight 0s from decimal 0 and last eight 0s from decimal 0 so DDN is 255.0.0.0. The prefix notation is /8
What is a FIN segment?
A FIN bit is used to request a normal connection closing. If a segment of connection needs to be close normally, it sets the FIN bit.
What is the IPV6 header?
A header that is 40 octets long which as a version number field of 0110, the traffic class and the flow label fields
What is IPV6?
A new version of the Internet Protocol
Address Resolution
A process for determining a host's data link layer address if you know its IP address
Next Hop Router
A router to which another router forwards a packet in order to get the packet a step closer to reaching its destination host
Prefix notation
A way of representing masks. Gives the number of initial 1s in the mask
What if in prefix notation, a mask is /16? Express in DDN.
Binary form is 11111111 11111111 00000000 00000000. So DDN is 255.255.0.0
What if in prefix notation, a mask is /18
Binary form is 11111111 11111111 11000000 00000000. So DDN is 255.255.192.0
How do we write 128bit IPV6 addresses?
By using hexadecimal notation
Decision Cache
Cheating. Nonstandard and somewhat risky of listing recent routing decisions for IP addresses
Subnet part
Comes after the network part, the subnet part bit specify a particular subnet within the network
Host part
Constitutes of the remaining bits in the 32-bit IP address which specifies a particular host on the subnet
Why is decision caching dangerous?
Decision caching assumes that the next link it is forwarding to is available, but if the link is broken or the next router becomes unavailable, it results into non-optimal routing and the packets may be sent into a black hole
Why are routing table more complex than Ethernet switching tables? Give a detailed answer.
Ethernet switching tables are lookup tables that store MAC address and output port. When a frame is to be forwarded, Ethernet switch selects the row with destinations MAC address and port number sends the frame. Hence they do little work per frame. Routing tables consist of IP addresses of hosts connected and address of their own. Routers are organized in a mesh and hence there can be many rows with the same IP address. So it is more complicated sine it maintains all possible paths present.
What are 1-bit fields called?
Flag fields
Distinguish between four-way closes and abrupt resets.
Four way closes uses FIN and ACK bit, it's normal operation followed for closing connection and it is where acknowledgment follows segments. Abrupt reset only uses RST flag bit, it is used when a problem is encountered and RST is never followed by acknowledgement
Sequence Number
Gives segments its position in a stream of segments
Hierarchical
IP addresses are hierarchical, they usually consist of three parts that locate a host in progressively smaller parts of the Internet
Why is handling options the way that IPV4 is undesirable?
IPV4 handles options where each router must check each packet for options and examine them which adds processing time and therefore cost per packet
Why do we need IPV6? (Outgrowing IPV4)
IPV4 supports only 32 bit addresses but IPV6 supports 128 bits. The limited number of IPV4 addresses is not sufficient enough for all existing internet users but IPV6 can be more than sufficient.
When a network mask is applied to any IP addresses on the network, what is the result?
If the network mask is applied to an IP address then it will generate the network ID for the network. For example, if a network mask is 255.255.0.0 and the IP address is 128.171.17.13 then the result of masking is 128.171.0.0 which is the network part. The result implies that an IP address 128.171.17.13 is available on network ID 128.171.0.0
What do the 1s in a network mask correspond to in an IP address?
In network mask, the 1s correspond to the network part of an IP address. Making an IP address with network mask gives network ID
What do the 1s in a subnet mask correspond to in IP addresses?
In subnet mask, 1s correspond to the network part and subnet part. Masking an IP address with subnet mask gives the network and subnet part
What should a router do if it receives several packages going to the same destination IP address?
Instead of repeating the 3-step process for each packet, it caches the decision for IP address and use this decision for successive IP packets going to the same destination
Border Router
It connects different networks
What is the third row in IPV4?
It contains the Time to Live field and the protocol field
What is the first row in IPV4?
It contains the version field (0100), the header length, the DSCP, the ECN, and the total length of the packet
What is the second row in IPV4?
It has the information that the destination host uses to reassemble fragmented packets
Network part
It identifies the host's network on the Internet
Know the network part in an IP address 128.171.17.13
It is 128.171
What is the address resolution protocol?
It is a dynamic mapping protocol that maps a logical address (IP address) to a physical address (MAC address)
Why is it important for networking professionals to understand TCP?
It is important since it handles connection opening, closing, packet sequencing and acknowledgement handling so understanding TCP can help a professional to be well-versed with networking
What is an internal router?
It only connects different subnets within a network
What is a border router?
Its main job is to connect different networks such as ISP networks to corporate networks
What are the three parts of an IP address?
Network, subnet, and host
How does the last extension header before a UDP datagram indicate that the UDP datagram comes next?
No idea
Do routers have a row for each individual IP address?
No, they instead have a row which represent range of IP addresses which can either be a subnet or a network.
If you see 59 in the next header field of a header, what will follow this header?
Nothing
Internal Router
Only connects different subnets within a network
Why are sequence numbers good?
Segment numbers helps in receiving transport service to put the segments in the correct order
What are the fields in UDP segment?
Source port, destination port, data field
What are the fields in TCP segments?
Source port, destination port, sequence number, acknowledgment number and data field
Why can TCP handle long application messages?
TCP allows segmentation of long messages and also allow reassembling them making them able to handle very long application messages
What are Openings and Abrupt TCP closes?
TCP also permits another type of close which is an abrupt close. Whenever either side wishes to end a conversation, it can simply send a TCP reset segment
For what reason is TCP complex?
TCP is complex since it deals connection opening, closing, packet sequencing and acknowledgment handling. IP is designed to be simple so the more complicated transmission control tasks are left to TCP
What are TCP messages called?
TCP segments
What are IP Options problems?
The IPV4 allows options. There are several and may come in any order. Some are only read by the destination host. A lack of required order means that each router must look at every option to see if it applies which is time consuming
If the ACK bit is set, what other field must have a value?
The acknowledgement field
Which header is used by IPsec?
The encapsulating security payload (ESP) header
Routing
The forwarding of IP packets
What is usually the only extension header that routers have to consider?
The hop by hop options header
Why is the approach of using optional extension headers desirable?
The optional extension headers allow router to ignore subsequent extension headers. Reading few if any additional extension headers reduces routing decision time.
Interfaces
The router's equivalent of a network interface card; a port on a router that must be designed for the network which it connects. Or it can be an outlet through which an object communicates with the outside world. Ports.
Default Row
The row of a routing table that will be selected automatically if no other row matches its value is 0.0.0.0
What does a row represent?
The rows represent the routes for a range of IP addresses
What is IPV4?
The standard that governs most routers on the Internet and private internets
If someone says that a flag field is set, what does this mean?
There are six single bit fields in a TCP segment. If the value of the field is 1, the field is set, if it is 0 the field is not set. For example, if SYN bit is set, it will have a value 1 and it means that the segment is requesting a connection opening
What is the total length of an IP address?
Total is always 32 bits
What is the maximum application message size when UDP is used at the transport layer?
UDP can use 65,536 octets of application message.
What are UDP messages called?
UDP datagrams
Why can UDP not handle long application messages?
UDP protocol does not support segmentation and reassembly. It does not contain any fields in header relating to sequence or acknowledge numbers.
How would decision caching speed the routing decision for packets after the first one?
When a router receives a packet, it finds all the rows matching the destination value then it selects the best among all rows. But with decision caching, router first does the 3-step process to find the best route and caches the decision for successive packets.
Why is a reset segment not acknowledged?
When reset bit is used, connection is closed immediately. It is not available for any incoming packets, not even for acknowledgement and hence acknowledging is worthless
Do Ethernet switches have a row for each individual Ethernet address?
Yes, Ethernet switching tables maintain a look-up table in which each row has a MAC address of a host and an IP address of its own