KRKeegan

-The Basics: The internet backbone is made up of many different technologies. When a packet is sent across the Internet it often traverses through many different technologies, such as fiber, copper or wireless. Many of the different technologies have different requirements, some of them have a maximum packet size less than others. When a packet is sent that exceeds the size of the maximum on the line the packet must be broken, fragmented, into multiple pieces before it can traverse the line. In the following example the packet travels across three networks with three different maximum sizes:?

Computer A sends the original 512 byte packet, the next gateway knows that the line it must cross can only handle 256 bytes maximum. The packet is split in half and sent to the next gateway. This gateway knows that the next line can only handle 128 bytes, so each 256 byte packet is split in half again. Computer B receives four packets that are individual fragments of the original 512 byte packet. Computer B is responsible for reassembling these fragments. Technically the administration of fragmented packets is handled in the header so that the router does not have to open up the entire packet to know the fragmentation status. Within the header there are three bits for the fragmentation flags. Their values and meanings are as follows:

Value	0	1
Bit 1	Reserved	Reserved
Bit 2	Allow	Don't Fragment
Bit 3	Last	More

So in the previous example in the network with 256B maximum the first packets flags would be 001 and the second packet would look like 000. Since the Internet is chaotic, packets that are sent first may not arrive first. This means that there must be a method to determine the correct order for the packets. This is accomplished by the offset value in the header. This 13 bit value contains the distance in bytes the packet is from the beginning of the original packet. To bring the entire concept together let us look at what the headers for the last part of the first example would look like:

Packet	Flags	Offset
1	001 - More Fragments	0000000000000 - Start at Byte 0
2	001 - More Fragments	0000010000000 - Start at Byte 128
3	001 - More Fragments	0000100000000 - Start at Byte 256
4	000 - Last Fragment	0000110000000 - Start at Byte 384

If the packets arrive out of order the buffer on the receiving computer places each packet at its offset value in memory. So that if all the packets but packet 2 arrive correctly the buffer memory will contain information from 0-127B and 256-512B. This also raises on major problem with fragmenting. Since fragmenting occurs on the line and not at the original computer when the destination fails to receive one of the fragments it must request the entire original packet again since TCP cannot communicate with the line. This will most likely result in the fragmentation of the packets again. Also since fragmentation occurs on the line it can actually slow down connection speeds. There are special technologies that enable you to determine the weakest link in the line and sets your maximum to this value so that fragmentation is not necessary.

[b]-The Basics:[/b] If Internet addressing remained as it was originally designed, it is likely that the number of hosts today would far exceed the available IP addresses. The problem with the standard classed addressing scheme is that the network ID length is predetermined by the class of the IP address. This also predetermines the number of hosts on each network. Below is a table of the standard classed IP address scheme:

Class	From IP	To IP	Maximum Networks	Maximum Hosts
A	1.0.0.0	126.255.255.255	126	1,677,216
B	128.0.0.0	191.255.255.255	16,384	65,536
C	192.0.0.0	223.255.255.255	2,097,152	256

It is obvious that a Class A IP range is unrealistic in its number of hosts. A network with this much traffic on the wire would crawl to a halt. This system does not allow more than a two tiered structure for IP addressing, this does not really match the actual layout of the Internet. Subnetting allows for differing lengths of network IDs. This means that Any IP class range can be divided up into multiple network IDs. The subnet mask essentially tells the router what part of the IP address to use as a network ID, as demonstrated in the following table using binary:

IP Address	10000000	00000001	0001	0001	00000001
Mask	11111111	11111111	1111	0000	00000000
Result	10000000	00000001	0001	N/A
Logical Address	128	1	1	N/A

The logical address is exactly what the router sees and it uses this to determine where to send the packet. This enables the CMC network to divide the Class B IP address range into class C subnets. This means that at a maximum there is only 253 hosts on one line. This subdivision enables added security within the network and organization of the network structure.

-Introduction:
For months I have been trying to create a Perl SMTP AUTH script that would allow me to log into my mail server, that requires authentication, from a perl script and send mail. I was under the impression that the encryption was CRAM-MD5, however I found that it other encryptions are acceptable and are much easier than this. I have tried this script on a number of servers and it appears to work well. I must thank Justin Clark for sparking my interest in this again.