Snow – A layer 3 virtual network that uses public keys instead of IP addresses

by fiatjaf • 11 years ago

444 92 11 years ago

Interesting, but I can't immediately think of any real-world use cases that this solves that aren't already solved with existing technology. The docs don't really describe anything beyond "automatic NAT traversal and end-to-end encryption with no configuration". No configuration? Not entirely true. :)

What are some use cases this can be applied to?

snissn• 11 years ago

i haven't looked into it deeply, but it sounds like it'd be great if you wanted to build a decentralized p2p system that didn't require a centralized routing system like dns or ip. like in practice email or twitter but decentralized, and would act similarly as simple as HTTP over DNS

deftnerd• 11 years ago

I'm trying to place my finger on it too but I feel an itch in the back of my brain that says "This is important and can be used for the core of something really amazing"

Isn't this a similar concept to Tor addresses without the onion routing being part of it?

baby• 11 years ago

in Tor servers are one IP and one public key, so yeah you can't talk to that ip if they change their public key. It's closely related. But this is because someone is giving you a table of that ip <-> public key association. If an attacker change the public key of an ip in that table then... That's why OP's idea is stronger.

woah• 11 years ago

When trying to design a decentralized Internet platform of any kind, one always runs into the issue that if it is built on IP addresses, it will be fundamentally centralized. This fixes that at a low level.

peterwwillis2• 11 years ago

First of all, No, a protocol or application is not fundamentally centralized because it uses IP. IP already supports multiple forms of addressing and routing and both centralized and decentralized services use IP.

Second, snow does not change whether an application is centralized or not. It's the application which is centralized, not the address. Your host's address can be "1.2.3.4" or "abcdefghijklmnop", this does not change how the application works at all.

Third, snow is just a tunnel. Any tunnel would "fix" an application the same way by simply translating addresses and encapsulating communication.

This is basically just onion routing, but snow doesn't really exist to be an onion router. The real purpose of snow appears to be that the author wanted to use the features of IPv6 (secure connections and the ability to address and connect to a host behind a network firewall) without having to actually use IPv6 in his application, and doing all this on top of an IPv6-only network. This is what sets it apart from every other NAT-tunnel. The public key stuff is a red herring.

politician2• 11 years ago

Off the top of my head, region-spanning container networking.

buss• 11 years ago

I had the same thought, but am curious if you've solved the following problems:

Would you bake the private key into the container or set it at runtime? If you set it at runtime how will two containers in different places know who to talk to?

Perhaps you generate the keys at build time and add the public keys to the partner containers, then at run time you inject the private key into the container via an env var. Now you have to securely manage and transport private keys and you've got two problems.

There must be other things I'm not considering.

And, of course, whatever system is running the container can step into it and read the private keys (or any malicious containers running on the host that are able to break out of the container). But we can just avoid that by saying they are our own hardware.

jacquesm• 11 years ago

Skype killer.

the84722• 11 years ago

I can't immediately think of any real-world use cases

The problem with SSL is that it needs certificates. You need a domain name and a certificate if you want to run anything over SSL in a resonable manner.

If address == identity then those requirements vanish because learning about the address already provides you all the information you need to establish a secure connection.

It democratizes authenticated connections.

marcosdumay• 11 years ago

So, your idea is to use a DNSSEC-like key distribution without a DNSSEC-like zone signing? Not convinced.

It's interesting, and it may be useful. But for securing what we currently do with IP it's useless.

lucb1e3• 11 years ago

Connecting two computers.

IP addresses (normal, unicast ones) like 1.2.3.4 and a9c::890 are meant to reach a computer somewhere on the Internet, through any number of routers. Routers are meant to forward IP packets until they reach their goal.

Try it with your neighbor's computer, it's not going to work. Did he enable DMZ or port forwarding? Alright, that works when your neighbor is home. Now try it when your neighbor is at work. His IP address changed, so there goes the reachability. This seems like "duh, obviously," and you're right. But I just want to perform the fundamental action of connecting two computers.

I personally used Tor to solve this problem: run a hidden service on one, connect to the .onion address on the other (you can configure ssh to work with .onion addresses).

This public key system would solve the problem in a much better way, without going through Tor. Not that Tor is bad, it's just not meant for this. Connecting the two machines directly without thinking about the intermediary network is what I wanted.

alistproducer22• 11 years ago

im still wondering how a intermediary dns-like server is not needed to perform the public key to ip address translation?

SomeoneWeird• 11 years ago

This seems very similar to cjdns[1]

[1] https://github.com/cjdelisle/cjdns

jMyles2• 11 years ago

CJDNS is amazing.

I've been working on a python frontend for it; I call it Cirque. https://github.com/jMyles/cirque

joeyh2• 11 years ago

Not realy. cjdns seems to need an invite to get onto its network.

Not saying that's necessarily a bad idea; cjdns seems to be useful to the people that use it. But if I want to build an app that communicates P2P over such a network, a manual step to join the network won't fly.

thescriptkiddie• 11 years ago

cjdns does not exactly need an invite. If you want to peer over the internet you'll need to swap public keys with someone already part of the network. But most people have turned on autopeering, which will discover peers on the local network and take care of peer swapping for you. Keep in mind that the intended purpose of cjdns is not to be a p2p library, it is to be a mechanism for connecting local mesh networks together into a mesh internet.

kpcyrd• 11 years ago

cjdns dev here, the network has been opened a while ago. There are some public peers floating around.

There's nightfall[0] to find/announce public peers, but I consider it very beta (quickly hacked together on a bus trip).

[0] https://github.com/kpcyrd/nightfall

gabeio• 11 years ago

cjdns can be used as a private mesh also so it completely depends on the application. A private mesh could be more appealing to some.

gesman• 11 years ago

ARP slow-ish?

But idea is well welcome. Key based address resolution has potential of obsoleting many types of current DNS attacks.

api• 11 years ago

No ARP. It's L3.

AaronFriel2• 11 years ago

This is similar to what I thought IPv6 IPsec should have been, auto-generated addresses: where address generation uses the hash of a public key. Sure, the addresses would have to be longer (in a /48, you only have 80 bits of choice), but if IPv6 were longer to accommodate strong hashes, it would solve much of the problem of secure computer-to-computer communication in a decentralized way.

Right now, IPsec practically requires PKI. But at Google or Amazon's scale, PKI is far from an easy problem, distributing keys to millions of nodes must be painful. And auditing the system must be its own level of hell, as I doubt many internal PKI systems attempt to manage devices at that scale. Unlike a smartphone or a laptop, where you can rely on 2-factor authentication, a server must be single-factor authenticated. The server is the server, and that places a huge burden on correctly allocating certificates.

And then there's the chicken and the egg problem: if you want to deploy PKI to millions of existing servers, how do you do that and ensure every server is what it says it is? There's too many shaky links of trust involved for a system like that to stand up.

I really like this idea, it's in many ways better than the idea I had about IPv6, because it uses the DNS layer to advertise public keys. It's inarguably more extensible, to boot. My idea would fix IPv6 into a single standard for IPsec, this is much more flexible.

jMyles2• 11 years ago

where address generation uses the hash of a public key.

This is precisely how CJDNS works, and it works wonderfully.

derefr• 11 years ago

I'm pretty sure this is already in IPv6: https://en.wikipedia.org/wiki/Secure_Neighbor_Discovery

AaronFriel2• 11 years ago

Hey that's pretty cool! I wasn't aware of this.

zurn• 11 years ago

Sounds like HIP (Host Identity Protocol).

exabrial• 11 years ago

Basically someone did what IPV6 should have.... Congrats!

alistproducer22• 11 years ago

maybe im missing something, but what is the point of using the public keys if they need to be mapped to ip? dont u still need dns servers to resolve the ip addresses?

purp2• 11 years ago

You get full encryption end to end, transparent to any app or program on the box, with next to zero config.

See http://trustiosity.com/snow/how-it-works.html

lucb1e3• 11 years ago

It uses DHT instead of DNS, but basically yes. The point, however, at least in my opinion, is this: https://news.ycombinator.com/item?id=9844987

gruez• 11 years ago

Wouldn't it be better if the address was the hash of the public key? That would significantly reduce the size of each address (great for UX).

y0ghur7_xxx• 11 years ago

Wouldn't it be better if the address was the hash of the public key?

It IS the hash of the public key...

-----[snipsnip]-----

This is a public key hash:

aaaac6rxjmmenb7m5txgpe3nmmrrh4z4ohcr7sxqkrvbk4csbrrorpw7.key

Snow lets you use it like an address...

-----[snipsnip]-----

kalmi10• 11 years ago

Yeah, no.. Gruez meant the actual IP(v6) address.

@gruez, The length of an IPv4 address would not allow for a future-proof-enough address length. IPv6 might just do it, I can't say.

@y0ghur7_xxx, Let's say I want to expose some snow-unaware service via snow to only a single host. I have no way to set up an iptables rule to do that atm. "When you resolve a key name an address is assigned to that key. The address remains assigned to the key as long as there is traffic, but never for less time than the TTL on the DNS record and never for less than 5 minutes (and generally for much longer than that)."

purp2• 11 years ago

A sweet hack, and full marks for humor in the FAQ.[1]

  Q: Is it secure? 
  A: Security is not binary.

  Q: OK, how secure is it? 
  A: It seems like you just asked that question.

  Q: No, the first question was if it's secure, 
     the second question was how secure is it. 
  A: Well now that wasn't even a question at all. 
     Tell you what, if you find an unreported security 
     vulnerability I'll buy you a beer.

[1] http://trustiosity.com/snow/faq.html

ytdht• 11 years ago

so it is as secure as every other pieces of software out there

cbd1984• 11 years ago

    Q: What attacks is it secure against?
    A: ...

    Q: How do you know it's secure against anything at all?
    A: ...

fallat3• 11 years ago

Personally I find this really unsettling for non-technical users. Almost asshole-ish. It's funny to us...but seriously providing a real answer after all that would even be sufficient. Not just answering with a "yes", but with a little blurb on how it is secure.

RyanZAG• 11 years ago

Oh, so when someone doesn't lie to you it's unsettling?

The only way to know if something is secure is when it's adopted en-mass and you see if it really was secure or not. You could read the WinXP pamphlet on security back when it was released and it had endless bullet points about how secure it was. It was probably the least secure software in the history of computing based on actual attacks after the fact.

Security isn't something you provide an answer to unless you're selling snake oil. Luckily, it seems most people prefer buying snake oil and are happy to eat up a vendor telling them how secure an utterly untested product is.

Security theory is not something you can understand as a non-technical user anyway.

tomtomtom777• 11 years ago

I am not sure how many non-technical users will have an urge to install something like Snow.

Swannie2• 11 years ago

Q: Is it secure? A: No.

Q: What do you mean "No"? A: We believe we have done a good job in securing it.

Q: So did you do a good job? A: We hope so!

Q: You "hope so", what sort of answer is that? A: Trust us. It's secure. We are not hackers. We don't want to steal your data. We did not put in any back doors. We audited the code ourselves. There are not any kernel level hacks, root kits, or otherwise. This has been tested against a variety of anti-virus scanners and none of them flagged anything. We're very good. Please please trust us?

noja• 11 years ago

Non-technical users are not reading github pages.

chrisdevereux• 11 years ago

Given that this is an unproven experiment that could potentially be misinterpreted as something more, that could a feature, not a bug.

BuildTheRobots• 11 years ago

Personally I find this really unsettling for non-technical users.

There is of course the counter argument, that if you're non-technical, you probably shouldn't be trying to implement a cryptographic layer-3 network for any reason other than "the lols".

boomshucka2• 11 years ago

Really, any user that doesn't "get" these answers shouldn't be anywhere near this.

perfTerm• 11 years ago

How many nontechnical users use github? Although there are plenty of developers with a minimal understanding of security as well

homulilly• 11 years ago

I don't think non-technical users will be using a project like this anyway.

0xdeadbeefbabe2• 11 years ago

That "is it secure" question is a lie in question form.

chm• 11 years ago

I doubt his software is intended for non-technical use.

kragen• 11 years ago

I’m not part of the Snow project, but I have the impression it’s still pretty experimental. If so, it’s probably better for non-technical users to remain unsettled about it for a while yet.

E1OX8• 11 years ago

It's funny and I get what he/she is trying to say, but it's not that clever because there's no definition of what a security vulnerability is.

Pond is a great example of doing this well: https://pond.imperialviolet.org/threat.html

"if an entity can do something that is not listed here then that should count as a break of Pond"

joeyh2• 11 years ago

I gave this a quick try, as I've been looking for something like this that works for several years.

Looks like the DHT used for NAT and resolving .key addresses is not currently online, at least my (very well connected) test machine wasn't able to connect to the 1 pre-seeded DHT peer.

Anyone gotten it to work outside of a single machine and ideally thru NAT?

lewisl9029• 11 years ago

Does this or CJDNS have any mechanisms in place to limit metadata exposure?

From what I could gather, both use public DHTs for routing, and AFAIK public DHTs in general can be rather trivially crawled for metadata.

The current generation Internet already offers plenty of methods to protect message contents, but very few can also obfuscate metadata, which can be just as revealing, but almost always much more readily accessible.

mirimir• 11 years ago

You can run this through a VPN with a port forwarded for UDP. And that VPN can be tunneled through a nested chain of VPNs. That provides some metadata obfuscation.

misterdata• 11 years ago

Interesting, and similar to Telehash [1]. Main benefit over Telehash seems to be that it can 'just work' with existing applications.

[1] http://telehash.org

higherpurpose• 11 years ago

Cool. Wouldn't it have been better to use Libsodium instead of OpenSSL, though?

Also a Rust version of this would be nice.

zrm6• 11 years ago

I'm very strongly considering replacing OpenSSL with NaCl/libsodium. OpenSSL is a nightmare.

jokoon• 11 years ago

So it's a DHT, so it's decentralized, pretty nice, but I still don't understand the purpose of this. Is this trying to improve security ? If so, how ?

What I'm more interested in, is a protocol that can let people share data on a DHT, which is resistant to denial of service and other security issues. I guess freenet is that already (somehow), but it's really not usable.

There are so many things in bitcoin I'd love to see in other standards, especially for messaging and forums. It would make things so much harder for the NSA and advertisers.

lucb1e3• 11 years ago

but I still don't understand the purpose of this. Is this trying to improve security ? If so, how ?

For my answer to this, see my comment elsewhere in this thread: https://news.ycombinator.com/item?id=9844987

fiatjaf• 11 years ago

GNUnet it is.

Pretty usable, has also a browser version: https://github.com/amatus/gnunet-web

Hyperborian• 11 years ago

Interesting. Reminds me of CJDNS:

https://en.wikipedia.org/wiki/Cjdns

qimfgh8ny7w• 11 years ago

Snow's author is also the user "zrm" here on HN.

https://news.ycombinator.com/item?id=8231026

rakoo2• 11 years ago

The main problem with using a public key as your identity is that once it is leaked, you no longer have an "identity" (or more exactly, it's not uniquely you). There is no way you can change your identification without changing your identity.

oleganza• 11 years ago

Multi-key addresses would be more robust. Plus, when it's an Open Asset on Bitcoin blockchain, then you can "rotate" keys by moving your identity into a different multisig address when one of the previous keys gets compromised.

dimman• 11 years ago

That's true, however you would still need to possess the private key to be able to validate.

nly3• 11 years ago

Depends if the network has memory. If you can publish a "this key is compromised" message then at least you can take down your own identity so the thief can't use it, and work on re-establishing a new one.

zrm6• 11 years ago

The main problem with using a public key as your identity is that once it is leaked, you no longer have an "identity" (or more exactly, it's not uniquely you). There is no way you can change your identification without changing your identity.

The main problem with using a public key as your identity is that it's a horrible string of gibberish that people can't remember. What you want is some way of mapping some friendly name to your key.

But you can resolve friendly names to keys however you like. You can update that mapping however you like. That is orthogonal to what snow does.

rakoo2• 11 years ago

True, you could have a bunch of SRVs in your DNS that redirects from _snow._tcp.my.best.friend.snow to aaa<..>.key or bbb<...>.key. You can even expire your keys after any amount of time for free thanks to DNS. Hey, that actually looks like a nice thing to add to snow !

sudioStudio64• 11 years ago

Wow. That's a really cool idea. I'm really interested in where this goes from a networking nerd point of view.

nly3• 11 years ago

So if i'm following this right, querying the builtin DNS server for a .key actually triggers DHT lookup and NAT setup? That's a cunning hack.

How are NAT entries recycled? Also couldn't SNI be used here to do this via a single IP?

zrm6• 11 years ago

So if i'm following this right, querying the builtin DNS server for a .key actually triggers DHT lookup and NAT setup?

That's the idea.

How are NAT entries recycled?

DNS responses have a TTL. The mappings last at least as long as the TTL and get extended if any traffic is sent to the address or there is another name lookup. After there is no traffic for a period of time the address goes back into the pool.

Also couldn't SNI be used here to do this via a single IP?

SNI is HTTP. Doing it this way works with other protocols too.

nly3• 11 years ago

SNI isn't just for HTTP. It's basically just a field in the TLS handshake that tells the server what name the client is interested in.

I can't think of a way to make this compatible with the DNS method you're using now though... you'd need a new address class that only ever returns a fixed IP by DNS, which was used exclusively for alternative determinations of the .key name requested. You could do this for TCP/80 with the HTTP Host header and TCP/443 with SNI, for instance. I'm wondering if one way to do it would be with haproxy to avoid having to implement this yourself.

Since a lot of connections going over Snow are going to be HTTP or HTTPS, this might make sense, at least for IPv4 where you IP space is limited.

cordite• 11 years ago

Is there a paper or something that describes how this works?

As fun as trudging through supposedly secure C++ code is, I'd rather have an understanding derived from the principles.

zrm6• 11 years ago

I can answer more questions later today.

The project is in the middle of a partial rewrite. The existing DHT has several issues and I'm replacing it.

The change is going to break compatibility, which made it into a much bigger change because it provided an opportunity to make several other compatibility-breaking changes. So I haven't been promoting the project recently and the DHT bootstrap node is currently offline.

There should be new code some time around the end of summer.

hyperion2010• 11 years ago

This is amazing. The ability to create private networks from arbitrary subsets of networked machines without any (serious) restrictions opens up all sorts of new possibilities. For some reason though the first use case that comes to mind for me is operating botnets.

Merkur• 11 years ago

seems cool, but what is it good for?

jfischl• 11 years ago

I think this is super cool but kinda bummed it is licensed under AGPL.

Snow – A layer 3 virtual network that uses public keys instead of IP addresses

Related Stories