1.1.d [iii] IPv4 and IPv6 fragmentation
One of the innovations introduced by IPv6 is the elimination of hop-by-hop packet fragmentation. With the new protocol, fragmentation is managed at the ends by means of a special extension header.
More specifically, there are two main differences: Difference one is the fields for handling fragmentation are not in the basic IPv6 header but are put into an extension header if fragmentation is required. This makes IPv6 fragmentation lean because this fragmentation extension header is only inserted in the packet, if fragmentation needs to be done. Difference two is that IPv6 routers do not fragment anymore. Fragmentation has to be done by the source host. He will evaluate the packet size by using Path MTU discovery.
In IPv4, fragmentation can be performed by a router between the source and destination of an IP datagram, but reassembly is only done by the destination device.
The IPv6 Next Header field is used to “chain together” the headers in an IPv6 datagram. The Next Header field in the main header contains the number of the first extension header; its Next Header contains the number of the second, and so forth. The last header in the datagram contains the number of the encapsulated protocol that begins the Data field.