News for hackers. Stuff that matters.
Re:An update on IPv6 by danellis on Monday June 15, 2015 @02:42AM (#9717858)
> lord is it wonderful not to have to play CIDR addressing games, everything is a ::/64

/64 is the smallest network, but there are certainly larger aggregates with smaller prefixes.

Re:An update on IPv6 by ghshephard on Monday June 15, 2015 @03:29AM (#9717982)
Yes - you're absolutely right. There are administrative collections of /32s, /48s, /56s - but when it comes time to lighting up a network interface - it's a /64. Even if I only have two devices on a network segment, (two routers that can perform ND with each other), a place that would have used a /30 in the IPv4 space, I'll light up a /64 and burn the other (2^64)-2 addresses. With the possible exception of some security scenarios in which a /127 might be desirable, there is no "address conservation" requirement to use blocks smaller than a /64.

For those IPv4 network engineers who have a decade+ of jugglings /25s, /26s, /27s, /28s and /29s, and trying to preserve IP space (particularly globally routable), this is nice change of affairs.

Re:An update on IPv6 by danellis on Monday June 15, 2015 @09:57AM (#9719279)
Okay, sure, on hosts you're only going to see /64s, but unless I'm seriously misunderstanding something, anyone dealing with route aggregation is going to have to deal with smaller prefixes.

And even just talking about "this is global unicast space", "this is 6to4 space" and "this is link local space" involves small prefixes. So, for example, you still need to do the same mental juggling to know that 2000::/3 is 2000... to 3fff....

Re:An update on IPv6 by ghshephard on Monday June 15, 2015 @12:48PM (#9720513)
Agreed. You can't get away from network blocks at the administrative routing level. But, at least life is a bit easier by making more common use of /48s and /32s - I.E. 2001:1868::/32 being 2001:1868:0000:0000::/64 -> 2001:1868:FFFF:FFFF::/64.

Thankfully, the 100x more common task of configuring an interface, will always be a /64.

Re:An update on IPv6 by kordless on Monday June 15, 2015 @02:50AM (#9717876)
Honestly, did he say otherwise? "Everything" in his world is "I use /64s everywhere."
Re:An update on IPv6 by danellis on Monday June 15, 2015 @02:57AM (#9717901)
"I use /64s everywhere" would be unusual for "someone who has worked on million+ IPv6 node deployments".
Re:An update on IPv6 by plq on Monday June 15, 2015 @02:53AM (#9717886)
I think he's talking from an end-user point of view.
Re:An update on IPv6 by danellis on Monday June 15, 2015 @02:54AM (#9717892)
Even from an end user point of view, the recommendation is to assign a /64 for a single network or a /48 for multiple networks. (Assuming 'end user = homeowner' here.)
Re:An update on IPv6 by ghshephard on Monday June 15, 2015 @09:11AM (#9719055)
And, as it turns out, the original recommendation to go with /48s turned out to be insufficiently flexible, and was overridden with https://tools.ietf.org/html/rfc6177.

In particular, some ISPs will hand out a /56 - allowing the home user 256 /64s. But, once again, these are administrative routing decisions - on the interface, it's always a /64 or (in some rare cases - though I'd love to hear if people still see this in the wild) /127s for router-router connections.

You will never, ever see, for example, a /72 IPv6 network in the wild.