Dan Kaminsky Suggests Having Fun with DNS

boogahsmalls writes "A few weekends ago Dan Kaminsky of scanrand fame presented some pretty cool ideas involving DNS that made plenty of heads spin at the LayerOne Technology Conference. Some of his concepts included Voice over DNS and storing Knoppix in a DNS cache. He's also apparently got a couple new tools in the pipe including a scanrand based DNS scanner and a visualization suite. Could another version of Paketto Keiretsu be in the works?" (OpenOffice.org does a great job of opening the PowerPoint slideshow.)

Comments

No thanks,

[Anonymous Coward]

I'd rather my dns just work.

Nice ideas

[Anonymous Coward]

but who doesn't have Knoppix in the DNS cache already anyway? Welcome to the 21st century buddy.

Re:Nice ideas

[Vihai]

In soviet russia, Knoppix has a DNS cache... oh nevermind....

use the DNS to store presentations

[Anonymous Coward]

I'd rather read his slides in binary from IN A records than open powerpoint.

RTFPP?

[Nethead]

Now we have to Read The Fsckin' Power Point?

Re:RTFPP?

[MisanthropicProgram]

Why is the parent flambait?

He does have a point.

Great Article

[Anonymous Coward]

It's a pity most of the slashdot crowd won't understand any of its technical merits at all.
Mark this as flamebait if you will, but come back in a while and read the comments, I promise there will be hardly any discussion of the paper.

Dan is obviously a very smart guy, I like his ideas about using http tunnel (it's a great program), I'm going to have to give some of these ideas a work out!

Bob

Re:Great Article

[wwest4]

The presentation is intriguing, but like any typical slideshow, lacking in specifics (things like "stuff=cool" aren't terribly telling). Unless you already know the DNS pretty well, it would be hard to infer the nitty-gritty of the talk from this ppt without thinking pretty hard about it, and you shouldn't fault a diverse group of geeks from different nerd realms for not being DNS power users.

Re:Great Article

[Anonymous Coward]

No, I guess I shouldn't. That was kind of elitest of me and I apologise. It's just frustrating sometimes to see a really good article on slashdot, digging in to hopefully read some good comments about it, and finding people can only post "humourous" stuff or other equally lame stuff. If I don't understand an article, I don't post on it.

You're also right about the powerpoint, it would have obviously been much better for us if we'd been there to hear his presentation. It still gives us a good insight to his ideas though.

Bob

Re:Great Article

[wwest4]

One thing that is cool about /. is that if you're willing to dig for a bit, there are some crazy-smart people who know the material. There are plenty sympathetic to your lament also.

Re:Great Article

[magefile]

I'd suggest Open Office. If you're on a dialup, and don't want to install several hundred megs, then look at the google cache - it'll have an HTML-ized version.

Re:Great Article

[jovetoo]

His techniques allow someone to set up a cryptographically secure network that most likely completely ignores firewalls. It features high bandwidth-high latency connection, low bandwidth-low latency connections and is virtually untraceable, even to both parties involved in the connection. An initial hostname and time would act as the 'phonenumber'. (By keeping a certain request alive, one can even implement a dailing service with TTL delay.) A message service is freely included.

It is virtually impossible to shut these networks down without replacing/patching dns. Not an easy task.
The bandwidth available to this network most likely exceeds that of most irc-botnets. Especially since the root servers are defending themselves against DDoS attacks.

The tools he's still developing might be able to trace these things but it will still require cooperation of dns server administrators (to get their logs). You will never get them all and you'll have a LOT data to process. Accorfing to this the ICS root server continuosly handles almost 8Mbps (and can handle upto 80Mbps) of traffic. I seriously doubt they can log that... (if so, transferring the logs would continually consume a healthy percent of the servers bandwidth.)

Pretty smart man indeed and very idealistic or shortsighted. Both the right and the wrong sort of people would pay a lot of money for that...

Re:Great Article

[Anonymous Coward]

Google Cache does ppt -> HTML; for this one, however, note that both text and background are white, so you need to select all to see the text.

I don't have PowerPoint here either... Or OO.o.

Re:Great Article

[rasz]

Dan is obviously a very smart guy

.. and copied DNS and other ideas from others.

Re:Great Article

[Glamdrlng]

I'd discuss the paper, but it's in a format I can't view.

Since you apparently lack the bare minimum of resourcefulness necessary to read the file, I'm sure it's our loss that you can't participate in the conversation.

Re:Great Article

[Glamdrlng]

Merits? The guy is proposing a system for conducting conference calls through firewalls by hijacking DNS servers, and you can use the term "merits"?

What you're overlooking is, if Dan could have these ideas, so could someone else. By sharing his ideas publically, he's giving whitehats and blackhats a level playing field.

Consider also, many common auditing tools were once considered blackhat programs. For example, If Mr. Kaminsky had written scanrand in the late 90's / early 2000's, back when port scanning was considered an invasive hacking activity by most, it would have gotten the same treatment. Personally, I think we should thank him for sharing his ideas instead of using them against us.

Re:Great Article

[Effugas]

Freaking Zalewski :-) I hadn't seen this paper. Super cool, it'll help the next version of this speech greatly!

(I directly name Zalewski in one of my apps; believe me, if I had seen this, I'd have credited him.)

--Dan

Search Service

[OzPhIsH]

Gee, maybe they could make the results of any unresolved queries forward users to a handy search page, instead of returning an appropriate 'not found' response!

Another pointless piece of information:

[YouGotServed]

Microsoft Powerpoint also does a great job of opening the PowerPoint slideshow.

Re:Another pointless piece of information:

[cgenman]

I can see where this is going:

1: Funny retort about clippy, modded +5 insightful
2: Serious post defending Power Point, modded -1 Flamebait
3: Humorous post about necessary height of a post to go over one's head, modded +2 interesting
4: Serious post questioning the connection between wooden posts and the stability of Microsoft Software, modded +2 Funny
Meta comment about the rediculousness of it all: Priceless.

Re:Another pointless piece of information:

[binux]

Meta comment about the rediculousness of it all: Priceless.

You missed - Post about spelling bee champ in tears.

Re:Another pointless piece of information:

[nathanh]

Meta comment about the rediculousness of it all: Priceless.

A Slashdotter who can't spell "ridiculous": inevitable.

Crazy!

[chill]

Most people are lucky if DNS just works without major headaches.

I could swear BIND and its config file is considered, along with Sendmail, one of the most convoluted programs in Internetdom. It, again along with Sendmail, is historically also one of the most bug-ridden and exploited.

And now someone is suggesting futzing around with it?! Why not just change your domain to "rootmeplease.com" and get it over with?

-Charles

Re:Crazy!

[Dwonis]

It's easy. Use djbdns for a little while. BIND stars to look very sendmail-esque after that.

Re:Crazy!

[MerlynEmrys67]

My favorite joke from years ago was

Q: What is the difference between a sendmail.conf file and modem noise
A:

Re:Crazy!

[Feyr]

i have both a djbdns server (for a customer, 1200 domains or so) and a couple of bind ones (~400 domains).

how the fuck can you say djbdns is easier than bind? if i want an A record in bind it's "IN A" (see, easily understood). if you want the same in djbdns it's some cryptic characters that make no sense (and is, of course, undocumented, or was last time i needed it).

now the best part. there's MULTIPLE characters to do nearly the same thing. if i recall a + is a straight A record, and a @ (i think) is an A+PTR

give me bind anytime, it's MUCH easier. though i'm about to move to powerdns or something with a mysql or ldap backend so customers can edit their zones easily

Re:Crazy!

[pyrrhonist]

And then hoping that you get it? The problem that there can be multiple spaces between the "IN" and "A"

And why exactly is this an issue?

grep -i 'in *a' file

Re:Crazy!

[mkettler]

I'm not sure which article it was, but perhaps it was referencing this study.

In it someone did phase-space analysis of the PRNGs used in DNS, and combined it with a birthday paradox style attack. In it, an attack on BIND 8 was shown to be 100% likely to succeed, BIND 9 20% and DJBDNS was 30%.

However, if you read the rest of the article, it points out that DJBDNS also uses a strongly random source port for the query, making it significantly more resistant to the attack, as the attacker would have to guess both the query ID and the source port simultaneously. (The two put together have about 1 billion possible combinations. The ID alone only has 64k.)

Unless there's some other DNS poisoning attack I'm unaware of, I think I'd prefer DJBDNS, as it's more resistant than bind 8 or bind 9, despite it's slightly less random output than bind 9.

(Note: bind 9 can be configured to use non-fixed query ports, but you'd need an kernel level random source-port patch to get good security out of this.)

Nasty Nasty HTML Version

[OverlordQ]

Enjoy

Note: Was converted with *gasp*powerpoint so yes it is horrible :)

SPF and SPF+ work over DNS

[ideut]

Dan isn't the first one to suggest novel new applications for the DNS. Many will also be familiar with SPF, the "spam permitted from" framework for defining permitted email senders. Microsoft have recently taken over the standard process and are proposing for the sender permission rules to be sent in XML format over DNS!

The open source community's response so far has been SPF+, which is essentially a technique of encoding the rules in TCL, which is served over DNS and executed on the mailserver. For obvious reasons, SPF+ will probably define the future of spam control on the internet.

Re:SPF and SPF+ work over DNS

[Effugas]

Hmmm. We've been hearing about agent technology / mobile code for years, and not only has its functionality been a bit sketchy at best, but its security is a nightmare. Note -- you can't post Javascript on Slashdot or PHP within common forums, and there's a reason.

Putting TCL in DNS as a commonly used standard is a bit worrisome -- you'd have programmatic access to an execution context within any mail server. Not rejecting the idea outright -- but what are the functionality gains that justify such an outright expansion of remote access to untrusted parties?

--Dan

Re:Win2k DNS

[MisanthropicProgram]

Or...maybe he really doesn't understand. You see, kiddo, those of us who've been in this business for a decade or longer know that you can't know everything. Those who say they know everything or are experts are mistaken or lying. This biz is just too large and diverse to know everything.

In other words kid, don't fuck with us old guys or we'll show you who knows shit!

Some of this stuff really makes alot of sense

[mcrbids]

Forget the current legal nightmare of this proposal - just roll with me...

This guy proposes putting content (eg Knoppix) into DNS.

Why is DNS particularly not well suited for this kind of distribution mechanism?

Seems to me that if the RIAA wanted to distribute their movies via broadband providers (an inevitability, I'm afraid) the biggest problem would be dealing with BANDWIDTH.

I always figured that ISPs would have to have some way to cache content locally so their Internet pipes don't get absolutely HAMMERED by all the people viewing the latest flick...

DNS already has a mature, stable, and lightweight caching mechanism in place. Why not use it?

Honestly, caching content a la DNS might provide a MUCH more efficient content distribution mechanism than, say, BitTorrent.

Where's the bad part of this idea?

Re:Some of this stuff really makes alot of sense

[markov_chain]

Content would probably get cached better with BT than DNS because of the dynamically constructed network topology. The caching in DNS works as well as it does because it happens along the domain name hierarchy (duh). The default topology probably wouldn't be very efficient for content.

Further, DNS would need to be upgraded. There is a good reason that short-term, experimental applications are better done at the ends; read the End-to-end arguments in system design for further insights.

Re:Some of this stuff really makes alot of sense

[kryptkpr]

Where's the bad part of this idea?

1) I think the requirement for caching sets of 4 byte IP addresses and 4 GB movies are quite different. Just because a system is good at one, doesn't mean it will automatically be good at the other. When I RTFA, the author made it quite clear that there was a 512-byte packet size limit, of which only around 50% could be useful for actual data. By the author's own estimation, it would take 35,000 DNS servers to host a single 700mb Knoppix image.

2) DNS is already an overloaded system, and his idea uses recursion, so it would place even more load on top of it.

If you think this is going to replace BitTorrent, you're off your rocker.

Re:Some of this stuff really makes alot of sense

[Bagheera]

Forget the current legal nightmare of this proposal - just roll with me...

Were that we could...

Why is DNS particularly not well suited for this kind of distribution mechanism?

Because DNS is designed to handle its hierarchical data, not massive amounts of content? The extra fields available in DNS are there fo, well, DNS related stuff.

Seems to me that if the RIAA wanted to distribute their movies via broadband providers (an inevitability, I'm afraid) the biggest problem would be dealing with BANDWIDTH.

I know you meant the MPAA, not the RIAA, but I think their biggest problem will be letting go of their deep seated need for control, rather than bandwidth. They can afford the pipe. And I, for one, would be incredibly pissed off to find the RIAA (or any other commercial service) caching their stuff on MY name server.

I always figured that ISPs would have to have some way to cache content locally so their Internet pipes don't get absolutely HAMMERED by all the people viewing the latest flick...

Like, say, USENET?

DNS already has a mature, stable, and lightweight caching mechanism in place. Why not use it?

We do. Millions of times a day. We use it every time we translate a name to an IP number. Looking up, say www.slashdot.org

Honestly, caching content a la DNS might provide a MUCH more efficient content distribution mechanism than, say, BitTorrent.

Highly unlikely. A highly effecient system dedicated to caching content will almost certainly be better than trying to do the same thing with DNS. It's simply not made for it.

Where's the bad part of this idea?

Inefficiency. Load on already stressed servers. Better existing solutions. Should I go on?

Dan's come up with some brilliant ideas over time. Definately A Geek's Geek. But this one sounds a lot more like one of his thought experiments than an actual proposal. Like directly burning CD's over an SSH tunnel...

Re:Some of this stuff really makes alot of sense

[strabo]

DNS already has a mature, stable, and lightweight caching mechanism in place. Why not use it?

What part of the word lightweight don't you understand?

Re:Some of this stuff really makes alot of sense

[Effugas]

It is indeed a thought experiment -- but one that's led to some interesting stuff. Voice over DNS was actually a really surprising hack -- here you have a globally deployed caching system, sometimes several levels deep, that actually has the capacity to host the minimal bitrate for a minimally compressed voice link.

There's millions of servers out there that we can interface with -- what's the impact of that? If nothing else, it's fun to be playing with something other than TCP headers :-)

--Dan

P.S. A broom can be used to sweep the floor -- or to knock something out of a tree, or to scare off a wild animal, or to burn for heat. There's something to be said for separating common uses from "inherent purposes". HTTP was certainly never designed to host as much dynamic content as it does now!

Re:Some of this stuff really makes alot of sense

[RAMMS EIN]

``HTTP was certainly never designed to host as much dynamic content as it does now!''

Nor was it intended to do sessions (think webmail), and it doesn't do a very good job at those. RPC over HTTP (useful for interactive applications) is even worse; the HTTP headers can easily outweigh the payload. A stateful protocol (like FTP) would be a better fit for those uses.

PDF Link

[kryptkpr]

PDF Conversion of powerpoint presentation

On my ISP's very fast webspace, but please post mirrors in case they decide to pull the plug.

Where's the innovation?

[Have Blue]

DNS is just a pervasive and well-organized caching broadcast protocol, isn't it? Right now, all it's been used to transmit is mappings of ASCII strings to IP addresses, and ancillary data related to that. Why is using it to transmit anything else particularly innovative? We didn't see this much enthusiasm when someone figured out how to send Knoppix over HTTP or Usenet.

Re:Where's the innovation?

[Effugas]

Putting data in DNS -- not new, I say that very early and often. What is sort of new is the idea that you can connect to many, many servers to amortize the download speed across Internet-scale networks, using their caches as short but useful term storage devices.

Also, short term caching allows for unexpectedly useful distributed voice transmission.

--Dan

Sticking Knoppix distro in a DNS cache....

[NemosomeN]

Discussed YEARS ago with the possibility to sticking the source of DeCSS into a DNS cache (Among other things). I would put the source in an HTML comment here, but alas, no comment tags.

Re:WTF Is This?

[EdMcMan]

Once reading the article you would understand.

If you put the presentation in DNS it would not be a problem.

anybody remember DNS MUDs?

[andrewagill]

You used to be able to play a text adventure game with DNS:

]$ nslookup - hastur.rlyeh.net
> set querytype=txt
> set domain=adventure
> 1

Alas, hastur has been down since around 1998, but you can still live the magic if you believe in yourself!

dangerous ideas, just think of akamai dns problems

[Anonymous Coward]

Dan's got some interesting ideas, I'll grant you. But considering how scanrand has toasted network equipment I've run it against in the past, I don't think I'm too keen on his take on this. The tunneling angle is interesting, but when he gets to content distribution - it starts to look like a DNS stress tester more than a useful application, and considering how akamai got hosed for a bit last week, I sure hope that not many people play around with Dan's ideas unless they have a clue as to what they're doing. Needing 35,000 servers to xfer 700MB's of data at a reasonable speed is NOT an interesting hack, but it sure sounds similar in some principles to a mass DDoS.

Yea baby!

[stienman]

Ok, so let's do this:

We've got the Kaminsky protocol connected to the
DNS protocol
the DNS protocol's connected to the
UDP protocol
The UDP protocol's connected to the
IP protocol
Oh hear the word of the inefficient!

The second verse is left as an exercise for the reader. Please keep in mind that writing another verse is somewhat more productive than implementing the aforementioned Kaminsky protocol.

-Adam

Re:Put up or shut up.

[Carnildo]

http://cr.yp.to/djbdns/guarantee.html

The djbdns security guarantee
I offer $500 to the first person to publicly report a verifiable security hole in the latest version of djbdns.

Examples of problems that do not qualify:

* Denial-of-service attacks. (BIND 9's fragility makes denial of service completely trivial; but an attacker can easily take down the Domain Name System without using any of BIND's bugs. The DNS architecture needs to be decentralized.)

Says it right there. It's a DoS attack that, by means of a series of specially-selected queries, forces worst-case behavior out of the caching algorithm.

Re:Win2k DNS

[silas_moeckel]

OK this is pretty OT as well but I'll have to agree to many people have no depth. But in reviewing a canidate it's generaly better to try and figure out how quickly they can get some depth. And knowing a little bit of everything and being able to go deaper quickly can make you a great CTO :) or consultant (IE not a temp staffer being called a consultant)

Re:Win2k DNS

[mabinogi]

I honestly don't know either. But apparently DNS is hard, even when you're using W2K.
I've never figured out how one of our network people was able to ACCIDENTLY add an NS record for one of our web servers instead of an A record, and I've definitely never figured out how it is that they couldn't understand what the problem was or how to fix it. They use Win2K on the DNS servers.

If it'd been Bind, they wouldn't have made the mistake in the first place, because there is no way you would accidently type "NS" instead of "A". Not to mention the fact that they probably wouldn't have attempted to make the change, and would have waited until the person who knew what he was doing was back.

I'm assuming that the person in question randomly clicked stuff until he had somewhere he could put a server name in....

Slides 1-10 of 44, and /.'s lameness filter sucks

[leonbrooks]

This paragraph is random crap to keep that fscking lame slash lameness filter happy, please ignore it. Getting this past TFLSLF was five times harder than copy-pasting the individual text elements. This paragraph is random crap to keep TFLSLF happy, please ignore it. Getting this past TFLSLF was five times harder than copy-pasting the individual text elements. This paragraph is random crap to keep TFLSLF happy, please ignore it. Getting this past TFLSLF was five times harder than copy-pasting the individual text elements. This paragraph is random crap to keep TFLSLF happy, please ignore it. Getting this past TFLSLF was five times harder than copy-pasting the individual text elements. This paragraph is random crap to keep TFLSLF happy, please ignore it. Getting this past TFLSLF was five times harder than copy-pasting the individual text elements. This paragraph is random crap to keep TFLSLF happy, please ignore it. Getting this past TFLSLF was five times harder than copy-pasting the individual text elements. This paragraph is random crap to keep TFLSLF happy, please ignore it. Getting this past TFLSLF was five times harder than copy-pasting the individual text elements.

This paragraph is random crap to keep TFLSLF happy, please ignore it. Getting this past TFLSLF was five times harder than copy-pasting the individual text elements. This paragraph is random crap to keep TFLSLF happy, please ignore it. Getting this past TFLSLF was five times harder than copy-pasting the individual text elements. This paragraph is random crap to keep TFLSLF happy, please ignore it. Getting this past TFLSLF was five times harder than copy-pasting the individual text elements. This paragraph is random crap to keep TFLSLF happy, please ignore it. Getting this past TFLSLF was five times harder than copy-pasting the individual text elements. This paragraph is random crap to keep TFLSLF happy, please ignore it. Getting this past TFLSLF was five times harder than copy-pasting the individual text elements. This paragraph is random crap to keep TFLSLF happy, please ignore it. Getting this past TFLSLF was five times harder than copy-pasting the individual text elements. This paragraph is random crap to keep TFLSLF happy, please ignore it. Getting this past TFLSLF was five times harder than copy-pasting the individual text elements.

Black Ops 2004 @ LayerOne

Dan Kaminsky

===page===break===

Introduction

• Who am I?
  • Senior Security Consultant, Avaya Enterprise Security Practice
  • Author of "Paketto Keiretsu", a collection of advanced TCP/IP manipulation tools
  • Speaker at Black Hat Briefings
  • Black Ops of TCP/IP series
  • Gateway Cryptography w/ OpenSSH
  • Protocol Geek

===page===break===

What's On The Plate for Today?
/* char descrip[256] = "You'll see"; */

===page===break===

What is DNS

• DNS: Domain Name System
  • Mechanism for translating human-readable names into machine routable addresses
• "Like 411 for the Internet"
  • As 411 usually but not always yields simple phone numbers, DNS usually but not always yields IP addresses
  • A: Given name, find IP
  • MX: Given name, find Mail
  • PTR: Given IP, find name
  • TXT: Given name, find "stuff"

===page===break===

"Useful" Traits of DNS
(Very Very Abridged)

• Hierarchical
  • .com says where to find addresses in .doxpara.com, and .doxpara.com says where to find addresses in foo.doxpara.com
• Recursive vs. Iterative Lookups
  • Iterative Lookup: Ask a server a question, it tells you where to go to find out the answer
  • Recursive Lookup: Ask a server, it goes out and finds out the answer for you, and tells you
  • It queries the hierarchy - which you may control

Parent is a troll linking to a troll

[jensend]

If you read the linked email and the replies to it, you will find that the linked post is a troll. For real information about SPF, visit spf.pobox.com.

Re:You have no idea how appropriate this is

[Effugas]

Well, there are two kinds of people in the world -- those who see SOCKS over SSH over TCP over HTTP over DNS over UDP as neat, and those who don't.

The DNS backchannel through a firewall, by abusing the heirarchy, is a real problem.

--Dan

Re:You have no idea how appropriate this is

[dlb]

Weird bionic encapsulations are 'neat' until you're the one trying to justify the bandwidth bill.

It's neat until you've gone into the next higher pricing bracket because someone decided to piggyback a bunch of other protocols on top of dns to your external name servers. Aside from breaking rfc, or causing a self-inflicted DOS, there isn't much you can do about it.
(On the other hand, this is a prime example why allowing recursive DNS requests externally is a bad idea.)

What I think is neat is stuff that's going to save me bandwidth, not increase freeloader traffic.

"DNS backchannel through the firewall" is addressed by sensible design and a good security policy.
Wrapping a server around an enforcement point like you described in your presentation is horrible design; any nutcase that implements that solution deserves problems.

~dlb