Broadband Tech > IPv6 > v6 subnetting

v6 subnetting

My best recommendation is to start taking the free IPv6 course from Hurricane Electric. The concepts of IPv6 are quite different from IPv4 because so many addresses are available.

»ipv6.he.net/

See the link at the top center of the page, "Try our free certification program!"

Tim
--
"Life is like this long line, except at the end there ain't no merry-go-round." - Arthur on The King of Queens
~ Project Hope ~

thank you for replying and for the suggestion tim.

I ve seen the link but it as such does not give information on subnetting, which i'm looking for. indeed, have registered in the site and will keep doing that.

but for now I m still looking answer to my question.

If you can help me understand v6 subnetting for the example, it would be gr8!!!

genseek

»www.ipv6calculator.net/

Above is a pretty decent site for helping with the math. Ultimately, if you want to fully understand how the subnetting prefix system works, you need to break out the math of the slash notation. It follows the same general principles of IPv4.

2340:1111::/32 will have a range of:

2340:1111:0000:0000:0000:0000:0000:0000 - 2340:1111:ffff:ffff:ffff:ffff:ffff:ffff

reply to genseek

reply to genseek
Do you know how to subnet in IPv4 notation genseek ?

If you do, then the only difference you have to keep in mind is :

a) each IPv6 octet is 16bits instead of 8 bits for IPv4.
b) IPv6 address length is 128bits instead of 32bits for IPv4.

Clever_Proxy gave you the address range, so you'd have 2^96 hosts, including
the network and broadcast address; 2^96-1 if you're not counting these two addresses.

Regards

reply to genseek

reply to HELLFIRE
You misunderstand what an octet is. It comes from the prefix "oct," meaning of course 8. It is precisely 8 bits (not 16). It was a name chosen to be architecture neutral. They're not the same as bytes, because bytes can have parity or ECC bits associated with them, making them 9, 10, or some other bits wide (think RAM modules and controllers). "Words" have had several (bit) widths, but most commonly a power of 2 which is also some multiple of 8.

You're confusing "octet" with "some number in some base between separators." In the case of IPv4, they're octets separated by ".". In IPv6, they're 16 bit words in hexadecimal separated by ":".
--
English is a difficult enough language to interpret correctly when its rules are followed, let alone when a writer chooses not to follow those rules.

Jeopardy! replies and randomcaps REALLY suck!

reply to genseek
I have found TLDP's HOWTO Linux+IPv6, especially section 2.3 through chapter 3 helpful for understanding addressing and subnetting. I would say you really should understand both, not just subnetting.

The key thing to remember is that IPv6 attempts to aggregate better than IPv4 so that the routing tables are more manageable. Another key to remember is the subnet used mostly for "non-routers" is /64 to accommodate stateless address autoconfiguration for link local addresses and in combination with router advertisements for global addresses.
--
English is a difficult enough language to interpret correctly when its rules are followed, let alone when a writer chooses not to follow those rules.

Jeopardy! replies and randomcaps REALLY suck!

reply to rchandra
@ rchandra
Point taken about 'octet'... I'm just using that as it's a term i'm comfortable with.

Tellya what, you find the official name for the 4-character 'chunk' of an IPv6 address between two ::'s and you
can post it up here. I'm sure we're all for learning something new every day.

Regards

I don't know of an official word. Considering that "octet" refers to an 8-bit (octal) segment rendered in decimal format, and IPv6 is a 16-bit segment rendered in hexadecimal, I would propose a mouthful of: "hexadecimet".

I swear I just thought it out without research! But:

»www.ietf.org/mail-archive/web/ip···871.html

--
Norman
~Oh Lord, why have you come
~To Konnyu, with the Lion and the Drum

Hexadecimet is a mouthful, however, like all things IPv6, it's much longer and verbose than it's predecessor.

Welcome to the new more verbose internet :P

reply to NormanS
I've always heard the segments referred to as "words" or "Chunks". Does anyone know if the IETF ever settled on a formal term?

RFC 4291 defines the IPv6 addressing scheme, and the authors refer to the field as the "16-bit pieces of the address" twice, and "16-bit chunk of the address" once. I think they were purposely being careful to use generic terms rather than invent some new buzzword.

Based on the computer science definition, I believe the correct term for a 16-bit value is "word". Sixteen bits = two bytes = one word.

...works for me Wily_One

reply to Wily_One
But is the 8-bit value of a dotted quad number a, "byte", or an, "octet"? Consider the context.
--
Norman
~Oh Lord, why have you come
~To Konnyu, with the Lion and the Drum

reply to HELLFIRE

reply to NormanS

Really? Synonymous I mean. I have only encountered "octet" in the context of the IPv4 IP address. In which sense, all "octets" are "bytes", but not all "bytes" are "octets".
--
Norman
~Oh Lord, why have you come
~To Konnyu, with the Lion and the Drum

Historically a byte was however many bits the hardware would take in (bite) at once. It was only over time that the meaning of 1 byte = 8 bits became standardized.

An octet unambiguously specifies 8 bits.

Wikipedia credits IBM engineer Werner Buchholz for coining the term byte in 1956.
--
Got some spare cpu cycles ? Join Team Helix or Team Starfire!