D-Link xStack DGS-3426P User Manual - Address Format

xStack DGS-3400 Series Layer 2 Gigabit Ethernet Managed Switch 

Extension Headers 

Extension headers are used to identify optional parameters regarding IPv6 packets such as routing, fragmentation of packets or 

authentication parameters. The types of extension headers supported are Hop-by-Hop, Routing, Fragment, Destination Options, 

Authentication and Encapsulating Security Payload. These extension headers are placed between the IPv6 packet header and the 

payload and are linked together by the aforementioned Next Header, as shown below. 

IPv6 header 

Next Header = TCP  

TCP header + data 

IPv6 header 

Next Header = Routing 

Routing Header 

Next Header = TCP 

TCP header + data 

IPv6 header 

Next Header = 

Destination Options  

Destination Options 

Header 

Next Header = Routing 

Routing Header 

Next Header = TCP 

TCP header + data 

Each header has a specific place in the header chain and must follow the following order: 

•  IPv6 Header 

•  Hop-By-Hop Header (Must follow the IPv6 header) 

•  Destination Options  

•  Routing Header 

•  Fragment Header 

•  Authentication Header 

•  Encapsulating Security Payload Header 

•  Destination Options Header 

•  Upper Layer Header 

There may be zero, one or more extension headers in the IPv6 header, they must be processed in order and they are to be in 

increments of 8 octets in the IPv6 packet. Nodes that do not recognize the field of the extension header will discard the packet and 

send a relevant ICMPv6 message back to the source.  

Packet Fragmentation 

At times, packets are sent out to a destination that exceed the size of the Path MTU, so the source node is required to split these 

packets into fragments in individual packets which will be rebuilt when it reaches its final destination. Each of the packets that 

will be fragmented is given an Identification value, by the source node. It is essential that each of these Identification values is 

different than any other fragmented packet recently sent that include the same source and destination address. The original packet 

is divided into two parts, a fragmentable part and an unfragmentable part. The unfragemntable part of the packet consists of the 

IPv6 header and any extension headers present, up to the routing extension header. The fragmentable part has the payload plus any 

extension headers that must be processed by the final destination node. This part will be divided into multiple packets that are of a 

size that can be accepted by the Path MTU. The IPv6 header is then included with this fragmented part and sent to its destination. 

Once all parts of the fragmented packet reach its destination, they are reassembled using the Fragment Identification value, 

provided that the source and destination addresses are identical. 

Address Format 

To address the problem of finding a larger pool of IP addresses for IPv6, the size and format of the IPv4 format needed to be 

changed. Quadrupling the size of the address, from 32 bits to 128 bits, and encoding addresses using the hexadecimal form were 

used to solve the problem. In IPv4, the format of the address looked like xxx.xxx.xxx.xxx, where the x’s represent integers from 

0-9 (ex. 136.145.225.121). Now in IPv6, the format of the address resembles xxxx:xxxx:xxxx:xxxx:xxxx:xxxx:xxxx:xxxx where a 

set of xxxx represents a 16-bit hexadecimal value (ex. 2D83:0C76:3140:0000:0000:020C:417A:3214). Although this address 

35