[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: current IPv6 implementation / configuration



Brian McCullough wrote:
> On Sat, Dec 27, 2008 at 06:55:50PM +0100, Jeroen Massar wrote:
>>
>>  = You know will most likely have at least a /64 for connecting to
>>    your ISP and if requested a /48 or /56 from them too.
> 
> Since my ISP ( as such ) doesn't seem to know anything about IPv6, I
> suspect that that assignment will be from HE or SixXS or somebody like
> that, correct?

Yep.

>>  * Draw a little map of your network
>>   - In that map everything on the same switch (L2, unrouted L3)
>>     lives in the same /64, thus out of your /48 assign one /64
>>     to that part.
> 
> I tend to think of my switches as hubs, since there is no way that I
> have to control them.
>
> In that case, do I just assign a flat network, or split it up as you
> have suggested, giving every switch a prefix and address as you
> recommend?  What I mean is that there is no external access to these
> devices, and therefore no method for assigning addresses internally.

Very simple test: make sure that all the hosts on your network are on,
then use 'ping6 -I eth0 ff02::1' to ping them all, you should get
replies from all the IPv6-enabled hosts on that subnet. The replies are
IPv6 Link-Locals though, which you can either match with the boxes in
question directly or convert them into MAC addresses (or check 'ip -6
nei sho' after pinging those addresses) and matching those up with IPv4
addresses or something similar. That way you know exactly what goes where.

Most very very likely your network is 'flat' though, thus all on the
same L2 portion. Just count the responses from the ping and you know how
many host there are on your network.

You can also of course just enable radvd for that subnet and check which
hosts get an IPv6 address from it :)

Many ways to figure it out, but generally you know that something is
routed. Wireless networks btw mostly are bridged.

Greets,
 Jeroen

Attachment: signature.asc
Description: OpenPGP digital signature


Reply to: