What IPv6 means for Bitcoin and the Blocksize Debate

[asperous]

I wanted to post this here in case anyone was interested or have some thoughts they'd be willing to share: https://www.reddit.com/r/Bitcoin/comments/3fdvx7/discussion_what_ipv6_means_for_bitcoin_and_the/

Here's my list for posterity:

[Con] Bigger headers
[Con] Possible Privacy issues
[Con] Possible Security issues
[Pro] DHCP not required (SLAAC may be used instead)
[Pro] Faster Headers
[Pro] Multicast Support
[Pro] NAT issues disapear, IPv6 Bitcoin nodes can accept incoming connections

[1714030913]

1714030913

[1714030913]

1714030913

[1714030913]

1714030913

[hexafraction]

As you mentioned in your other topic, multicast has some implementation issues. Also, can you state which exact security issues other than privacy are concerning given Bitcoin's existing trustless security model?

[asperous]

As you mentioned in your other topic, multicast has some implementation issues. Also, can you state which exact security issues other than privacy are concerning given Bitcoin's existing trustless security model?

Simple: Previously many computers were hidden behind a NAT and therefore couldn't receive any incoming connections. Once IPv6 rolls out people's computers will suddenly be routable. Now if a remote code execution 0day for some service you have running (like cups, dhcp, etc.) people can just spam the exploit across the net and collect private keys and install keyloggers to get passwords, etc. Even if you have a firewall most people just spam the "allow" button and enable incoming connections on all their programs.

[Schleicher]

We are running out of IPv4 addresses.
And now imagine millions of people in africa (for example) coming online.
They will have no choice but to use IPv6.

As far as firewalls are concerned:
good luck making them idiot proof and at the same time not annoying.
That's not going to happen any time soon.
I think educating people is the better option.

[vvic]

heh, IPv6 was first introduced in like 1992 or so... They were crying about running out of IPv4 back then already. IN reality there are tons of wasted IPs. Companies like apple have entire subnet of 17.0.0.0/8. Why do they need so many? https://en.wikipedia.org/wiki/List_of_assigned_/8_IPv4_address_blocks
and so on and so on. Another issue is NAT. NAT in the 90s was viewed as necessary evil, and they made sure NOT to make it work in IPv6 (although 5-8 years ago they sort of did make it work, also you can do NAT from IPv6 to IPv4 and vice versa) today however NATis the most basic and cheapest way to get security at home. Imagine all those home user's computers will become visible from the internet... Heh, no. NAT isn't going away even with IPv6. Many companies are not even looking into upgrading to IPv6. What will happen is many of them will keel use internally IPv4 and just have external IPv6. So multicast fancyness and other IPv6 features won't be seen in many networks simply because only edge router will have only external IPv6. So I'm not even sure how will bitcoin benefit from IPv6 as of right now. However,  I have bigger concert about block size... Few weeks ago I had bad storm coming through the area, killing power for couple hours, UPS only lasts 90min. I wasn't home, so my bitcoin-qt machine eventually lost power, which crashed my wallet badly. To the point it wouldn't load wallet.dat and blockchain db. well, of course I have wallet.dat backed up, but I didn't backup blockchain... I though why bother? I can always re-download it.  Well, it took me 8 days to download block chain and the size is 49.5gb or so. Now I'm backing up blockchain also. What will we do when it reaches 100-500gb?

[jl2012]

Well, it took me 8 days to download block chain and the size is 49.5gb or so. Now I'm backing up blockchain also. What will we do when it reaches 100-500gb?

If you can't afford to keep the whole blockchain, you can run in pruning mode so you only need to backup about 2GB of data