Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

and whats stopping an intruder reassigning that block to something that can be publically accessed and how will that be monitored?


Uhhh... Whut?

You obviously need to use a ULA prefix, they are not routed (just like RFC1918 space in IPv4).

Or you can just use your allocated IPv6 space, and firewall at the border. And you hopefully have BGP hijack monitoring set up anyway.


you do know single network devices can have more than one ip address?

afaics, the biggest issue with ipv6 is if its active all devices on a network can easily be coaxed to never route traffic anywhere near the router/firewall the network admistrator intended, simply by handing out extra routing info for alternate networks.


> afaics, the biggest issue with ipv6 is if its active all devices on a network can easily be coaxed to never route traffic anywhere near the router/firewall the network admistrator intended, simply by handing out extra routing info for alternate networks.

This is not unique to IPv6.

ARP spoofing is the v4 version of this attack. RA spoofing is the v6 version of the attack. In both cases, the solution is the same: lock down your L2 by enabling MAC / ARP / RA filtering on your switch.


true, but getting even a single public ipv4 address is hard.

anyone and everyone handing out public ipv6 addresses is by design.


I have 32 IPv4 addresses, how do I utilize them to hack Amazon?

It doesn't matter that you can get IPv6 addresses, you still need to be able to get onto the L2 network of your victim company to be able to mount RA attacks. You also will somehow need to force them to announce your IPv6 space to their peers.


with IPv4 you cant really, because getting traffic routed to those ips is a major undertaking.

with IPv6, every IPv6 capable device is potentially capable of handing out something like the entire IPv4 space of public ip addresses regardless of how a single firewall or router is configured.

"trying to configure connectivity and access resources using only IPv6 addresses is borderline insane"


You clearly don't understand how routing (and the Internet) works. My IPs are useless because I can't force the victim to route to them.

I similarly won't be able to force the victim to route their traffic to IPv6 addresses that I control.


what difference do you think it makes who controls the public ipv6 address.

with ipv6, they got one, all devices on the network are now by default accessible from the public internet instead of invisible to it . Thats the whole point of ipv6.


Reassigning a block sounds a lot harder than the IPv4 version of making something accessible, sending out a single packet to hole punch the NAT.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: