Difference between revisions of "IPv6"

From Hackepedia
Jump to navigationJump to search
m (RFC mention)
m (part of header)
Line 12: Line 12:
  
 
One good thing to remember is that an IPv6 packet header is on a 64 bit boundary.  So one has as part of the IPv6 header 64 bits data for things such as version and hop count and 128 bits source address and 128 bits destination address.  In a hex dump the addresses are easy to identify unlike in IPv4.
 
One good thing to remember is that an IPv6 packet header is on a 64 bit boundary.  So one has as part of the IPv6 header 64 bits data for things such as version and hop count and 128 bits source address and 128 bits destination address.  In a hex dump the addresses are easy to identify unlike in IPv4.
 +
 +
from RFC2460
 +
    +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
 +
  |Version| Traffic Class |          Flow Label                  |
 +
  +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
 +
  |        Payload Length        |  Next Header  |  Hop Limit  |
 +
  +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
 +
  |                                                              |
 +
  and 128 bits source and destination addresses..
 +
  
 
An IPv6 [[router]] never fragments, only the end-points in a session do if they wish.  Also unlike in IPv4 there is no checksum processing in the header and this should speed up routing.
 
An IPv6 [[router]] never fragments, only the end-points in a session do if they wish.  Also unlike in IPv4 there is no checksum processing in the header and this should speed up routing.

Revision as of 08:55, 1 April 2009

IPv6 (IP version 6, RFC 2460) is the replacement for the current IPv4 Internet. Why there is no version 5 is not known. The bonuses for IPv6 include 128 bit addresses instead of 32 in IPv4. This allows the Internet to grow so that every human on earth can have an IP address, as currently IPv4 only has space for only 4 billion IP addresses.

The IPv6 packet looks like this in a tcpdump (an ACK in a tcp exchange):

17:25:20.181441 2001:a60:f074::2.35081 > 2001:a60:f074::3.1022: . ack 22 win 16384   
<nop,nop,timestamp 1165806449 3659414783> [flowlabel 0x847fd] (len 32, hlim 64)
 0000: 6008 47fd 0020 0640 2001 0a60 f074 0000  `.Gý. .@ ..`ðt..
 0010: 0000 0000 0000 0002 2001 0a60 f074 0000  ........ ..`ðt..
 0020: 0000 0000 0000 0003 8909 03fe ad53 070a  ...........þ­S..
 0030: e04b ebca 8010 4000 cb84 0000 0101 080a  àKëÊ..@.Ë.......
 0040: 457c cb71 da1e 3cff                      E|ËqÚ.<ÿ

One good thing to remember is that an IPv6 packet header is on a 64 bit boundary. So one has as part of the IPv6 header 64 bits data for things such as version and hop count and 128 bits source address and 128 bits destination address. In a hex dump the addresses are easy to identify unlike in IPv4.

from RFC2460
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
  |Version| Traffic Class |           Flow Label                  |
  +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
  |         Payload Length        |  Next Header  |   Hop Limit   |
  +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
  |                                                               |
 and 128 bits source and destination addresses..


An IPv6 router never fragments, only the end-points in a session do if they wish. Also unlike in IPv4 there is no checksum processing in the header and this should speed up routing.