Two sides of the security coin

In February 2000, David Dittrich, the 39-year-old security administrator for the University of Washington, and "Mixter," a 22-year-old creator of tools for launching attacks against Web sites, faced off virtually when Mixter's program--the Tribe Flood Network--was used to inundate Yahoo, CNN and six other major Internet sites with data.

The denial-of-service attacks slowed access to the sites--and in some cases made the sites unreachable--for hours at a time.

Dittrich, who had analyzed TFN and other denial-of-service tools, became the expert of the hour, while Mixter--because the culprit who used his tools hadn't been found--became its villain.

In a recent interview at the CanSecWest conference in Vancouver, British Columbia, the two revealed that though they use their skills differently, Dittrich knows a lot about hacking and Mixter is well-informed about security.

When did you start doing security?
Dittrich: I actually sort of grew into it as a by-product of doing support. I taught myself, then started with the University of Washington doing support for the Unix workstations, and there were so many Unix compromises that I had to end up helping people figure out what happened and how to secure their systems. And it was such fascinating stuff.

When did you first start seeing the denial-of-service attacks? Your systems were being used to launch the attacks against the University of Minnesota, right?
Dittrich: Yeah, actually a little bit before that. We had DOS attacks going against our systems for years...It wasn't until May or June 1999 that we started seeing Trinoo (an early distributed denial-of-service tool) on a bunch of systems.

And then you saw the distributed denial-of-service (DDoS) attacks in February of 2000?
Dittrich: Yeah, against Yahoo. And that's the thing--everybody says DDoS, Feb. 8, that's when it happened. But no, it had been going on long before that.

Mixter, when did you start doing security?
Mixter: I have been interested in security for many years. In 1997 or '98, I was on (chat service) IRC and doing pretty much stupid things, and then basically the police showed up. Actually it was not that much of a problem. But after that I started seriously getting into coding and everything.

's="" systems="" on="" my="" own."="" --="" dittrich"=""> What did the police say when they showed up? Why did they show up?
Mixter: I did not do intrusions, but I used some German public phone numbers and then later I found out that I wasn't supposed to use them. At that time, I was not aware of the legal situation at all. I didn't even consider what I was doing to be illegal.

(Outside attack) was always a problem, because our network had no kind of protection. We were getting viruses and all kinds of things. I got to analyze an awful lot of intrusions.

Dittrich: Did you ever take programming classes?

Mixter: No, no, no. I started on my own. I was good at...machine language.

I went to college for a year and a half. I think I will do college when I finally have some success.

When you did create TFN?
Mixter: When I actually thought about it was in 1999, after all the other DOS (attacks) went down. I'd always try to go on IRC and talk to other people and use my reputation or whatever. I got enough information to relate to what the architecture has to look like. That's when I already knew something about router programming.

Are you still maintaining TFN?
Mixter: No. I am now programming for my company. It might become popular again, so I would rather not.

So do you consider yourself a black hat (a hacker who breaks into computers illegally)?
Mixter: No, not at all.

Gray hat? White hat?
Mixter: Actually, white. But I guess others think gray.

How about you (Dittrich); are you white hat (a hacker who works to improve system security)? I saw that you took a black hat at registration.
Dittrich: That's a functional reason. Anything white with me is going to end up dirty.

No, I used to do little prank things in college, but I never got into breaking into other people's systems on my own.

Did you care that TFN was being used to attack Web sites, Mixter?
Mixter: No, at the time I didn't care about the bandwidth attacks when I made the first version. The second version I made much more technical, so the script kiddies wouldn't be able to use it. But the first version I didn't very much want a technical challenge.

I think you really see a small group of people are causing a major amount of problems.

What did you think of TFN?
Dittrich: It wasn't used very long. Stacheldraht was much friendlier.

What are you working on now, Mixter?
Mixter: I am looking at something you could use to make worm packages, but with more intelligence than those today.

Dittrich: It seems that would be the next step. Automate the creation of back doors.

'we="" don't="" need="" take="" care="" of="" this="" until="" it="" becomes="" problem.'="" they'd="" much="" rather="" add="" new="" feature."="" --="" dittrich"=""> What do you think about people creating tools that can be used maliciously, like what Mixter did?
Dittrich: Well, there's a lot of riding on the backs of giants that goes on--taking advantage of people who know something that you can use. But there is a lot more noise out there that they create, like script kiddie attacks because they can now do it.

There is some truth in that a lot of these problems, no one sees as being a problem--like some of these vendors or customers. You go to a software company and they say, "We don't need to take care of this until it becomes a problem." They'd much rather add a new feature.

In a way, you almost need something to happen, but it's a waste of money.

So you need something like the Melissa virus, which didn't do a lot of damage, but raised awareness?
Dittrich: Yeah, if it had that effect, it is good. But I don't think it is viable to do that sort of thing for that point.

So in your mind it's a question of whether the ends justify the means?
Dittrich: I guess a better way to say it is if Mixter is out helping develop collusion, that's when I have a problem. It's hard, because look at the amount of time it takes to figure out what happened. It takes a lot of people to do a lot of work and a lot of skill.

Either apply it creating problems for other people or apply it to creating solutions. Because if all the people that were out there developing these thing were out there developing solutions...security solutions--and all the little kiddies that were out looking to emulate somebody were emulating that--then we wouldn't have so many programs that poke holes in products.

Mixter: There is no relation between people who write the exploits and the holes in the program. One point why I have given up exploit writing myself is that it didn't make a difference. If someone found a hole, within two or three days there would be at least one exploit written.

Do you regret writing the Tribe Flood Network?
Mixter: No. Maybe the way I published it could have been different. I only started doing professional security with the company in late 1999.

You are doing a lot of computer forensics stuff. Are you going to be doing that for a while?
Dittrich: Yeah, it is a really interesting area. There are a lot of problems to be solved. There aren't that many people in it who are really experienced. The administrators don't know how to use simple tools...They don't know how to think about low-level functions.

I can tell them, "Here is traffic coming from your system. I know your system is hacked." And they say, "Well, the computer doesn't show anything like that, so...you're lying to me."

Then they want me to tell them how they broke in, but you can't do that unless you have access to the system. And there are only so many people who do forensics. It is pretty challenging.

What do you think the biggest problem for security is right now?
Mixter: It's still education, right?

Dittrich: Yeah, probably. Most people don't understand either from the system owner or the user perspective how the security holes are there.

I keep using the analogy of the Wild West. When people first moved out here, they had no idea that they had to look out for bears and Indians running around. It was a pretty hostile environment that they moved into with whatever skills they had. A lot of people died, trains were being robbed all the time. It took a long time before the legal structure was in place or they could protect money transfers and stuff like that. That's where we kind of are right now.

A lot of people are going to suffer until they realize that the Internet is a more hostile environment than they first realized.

So you think that people should be educated before they get online?
Dittrich: I think that we should require that someone show a certain level of skill before we give them an IP address. There is a huge cost to not being on the network, so perhaps that will teach them. We have huge warning labels on hair driers; why not on the Internet?

When I went over to Europe, it really tripped me out. We went up this rail car to the top of the mountain. There was a family there having a picnic and there was a little kid dangling his feet over the edge. And my dad said, "What a terrible father. This kid's going to fall." I think you teach little kids to be careful around cliffs. You are responsible for yourself.

Do you think companies are standing next to the cliff?
Dittrich: I think companies are running toward the cliff. "Let's put everything on the Internet" is what they say.