IPv6

godofthunder9010

My first take of the CCNA (scored 7900) had a surprising number of IPv6 questions on it. I want to make sure I don't get burned again on that. I understand the general concept, but I don't know IPv6 very well. IPv6 is daunting. Each address is huge, and IPv6 has a bunch of functionality that I'm not used to thinking as "built into the address." I haven't really gotten into it at any great depth. Lovely that it does a lot of nifty tricks. Right now, I only care about passing the CCNA exam. I'll worry about learning to stop hating IPv6 later. I understand the need for it. I just despise the size of each address and the extreme complexity. It's just a lot more address and info to keep track of.

Okay, so a couple of things that IPv6 does that I'm confused about:
1.) One source says every IPv6 address contains a loopback address. Loopback as in 127.0.0.1 loopback? What's the point? Besides, IPv6 also has it's own loopback(0:0:0:0:0:0:0:1 aka ::1), so what is this referring to? Some other definition of "loopback"?
2.) IPv6 is supposed to put the MAC address directly into the address, but it's supposed to divide it by putting a 0xFFFF (I'm assuming thats just ":FFFF:") right in the center of the MAC. Is that correct? Why break the MAC in two anyways? From what I understand, Windows Vista implementation of IPv6 doesn't do this anyways. Anyone know more about this?
3.) What other things are testable on IPv6?

luke_bibby

790 is really really close to the passing score so don't feel discouraged.

1.) One source says every IPv6 address contains a loopback address. Loopback as in 127.0.0.1 loopback? What's the point? Besides, IPv6 also has it's own loopback(0:0:0:0:0:0:0:1 aka ::1), so what is this referring to? Some other definition of "loopback"?

Yes the IPv4 loopback address of 127.0.0.1 is the IPv6 equivalent of ::1. The loopback address is used to test the local TCP/IP stack.

2.) IPv6 is supposed to put the MAC address directly into the address, but it's supposed to divide it by putting a 0xFFFF (I'm assuming thats just ":FFFF:") right in the center of the MAC. Is that correct? Why break the MAC in two anyways? From what I understand, Windows Vista implementation of IPv6 doesn't do this anyways. Anyone know more about this?

When you use something like stateless autoconfiguration, you find the 64-bit subnet prefix from the local router using an NDP Router Solication/Router Advertisement message exchange. A MAC address is only 48-bits and the Interface ID (equivelant of the host ID in IPv4 addresses) is 64-bits wide, so the local host pads the MAC address by putting four hex characters in the middle - the hex characters are FFFE. It also checks that the 7th bit in the Interface ID - if it isn't, then it changes the 0 in the address to a 1. Remember that the aim of the game with this is to generate a unique 128-bit address on the local subnet.

This process is called EUI-64, and from memory its an IEEE standard not a Cisco proprietary thing. So it should be the same in Windows provided they are complying with the EUI-64 standard.

3.) What other things are testable on IPv6?
Check out the ICND2/CCNA exam blueprint on the Cisco website. These are the IPv6-specific topics:

• Describe the technological requirements for running IPv6 (including: protocols, dual stack, tunneling, etc)
• Describe IPv6 addresses

Hope that helped.

godofthunder9010

Yes it definitely helps and thank you.

The part about the loopback address puzzles me because it seems to imply that all IPv6 address have a built in loopback no matter what. Okay, so what that seems to tell me is that 2001:db8:85a3:0:0:8a2e:370:7334 has a loopback address built into it. And yet we also know that ::1 is THE loopback address for IPv6. That's where I'm confused. I know that the loopback tests the TCP/IP but I'm at a loss as to what to make of the statement "every IPv6 address contains a loopback."

I remember my CCNA course instructor mentioning that Vista doesn't comply with the standard of putting the MAC into the IPv6 address. I've not messed with IPv6 on Vista to verify of course, I'm just assuming he knew what he was talking about. If Microsoft isn't on board with the business of putting the MAC in the IP, then things are going to get confusing as implementation moves forward. I can think of a lot of reasons why EUI-64 would be a very good idea.

Thanks for the help.

godofthunder9010

luke_bibby wrote: »

• Describe the technological requirements for running IPv6 (including: protocols, dual stack, tunneling, etc)

What is the short and concise list of requirements to run IPv6 out of curiosity?

luke_bibby

Off the top of my head, you should know something about the following:

NDP
DHCPv6
Stateless autoconfiguration
EUI-64
RIPng
Teredo tunnels
Dual stacking

Probably more but just give the Lammle/Odom books a read over

mattrgee

I wrote an article about EUI-64 recently, pretty cool:

IPv6 EUI-64 Addressing | BitsOnTheLine