Thursday, December 16, 2010

Cracking hashes in the JavaScript cloud with Ravan

Password cracking and JavaScript are very rarely mentioned in the same sentence. JavaScript is a bad choice for the job due to two primary reasons - it cannot run continuously for long periods without freezing the browser and it is way slower than native code.

HTML5 takes care of the first problem with WebWorkers, now any website can start a background JavaScript thread that can run continuously without causing stability issues for the browser. That is one hurdle passed.

The second issue of speed is becoming less relevant with each passing day as the speed of JavaScript engines is increasing at a greater rate than the increase of system speed. It might surprise most people how fast JavaScript actually is, 100,000 MD5 hashes/sec on a i5 machine (Opera). Thats the best number I could get from my system, in most cases it would vary between 50,000 - 100,000 MD5 hashes/sec. This is still about 100-115 times slower than native code on the same machine but that's alright. What JavaScript lacks in outright speed can be more than made up for by its ability to distribute.

It is trivial to get someone to execute your JavaScript in their browsers, just get them to visit a link and you have remote code execution of the JavaScript kind, they don't have to download or install any applications on their system or have any special privileges. It is ridiculously easy to distribute computation with JavaScript. And with about 110 browsers pointed to your site you have already achieved the speed of native code on one machine. With 1100 browser that is equivalent to 10 machines cracking passwords in native code.

To demonstrate this I have built Ravan a JavaScript Distributed Computing System that can crack MD5, SHA1, SHA256, SHA512 hashes. Details on how it works and how to use it are available here. It was released at BlackHat Abu Dhabi last month and has already had over 700 hash submissions. Both the cracking of the hashes and management of the distribution process is done in JavaScript.

The commercial cloud might have made cracking hashes super cheap but the JavaScript cloud has made it free.

Wednesday, December 15, 2010

Performing DDoS attacks with HTML5 Cross Origin Requests & WebWorkers

Update: Shellex has performed detailed performance analysis of this technique.

DDoS attacks are the rage this year, atleast in the latter part of the year. There have been numerous instances of successful DDoS attacks just in the past few months. Some of the current DoS/DDoS options seem to be LOIC, HTTP POST DoS and Jester's unreleased XerXes.

This post is about a DDoS technique I spoke about at BlackHat Abu Dhabi that uses two HTML5 features - WebWorkers and Cross Origin Requests. It is a very simple yet effective technique - start a WebWorker that would fire multiple Cross Origin Requests at the target. This is possible since Cross Origin Requests that use the GET method can be sent to any website, the restriction is only on reading the response which is anyway not of interest in this case. Sending a cross domain GET request is nothing new, you can even do that by embedding a remote URL in the IMG or the SCRIPT tag but the interesting part here is performance. My tests on Safari and Chrome showed that both the browsers were able to send more than 10,000 Cross Origin Requests in one minute.

So simply by getting someone to visit a URL you can get them to send 10,000 HTTP requests/minute to a target of your choice. Now if you pick a juicy target URL, one that would make the server do some heavy processing then just 10,000 requests/minute might overwhelm it. Lets scale this a little, say 60 people visit the URL containing the DoS JavaScript, that is 10,000 requests/second at the target. With just 6000 visitors to this URL we can send around 1 million requests/second to the target. Getting 6000 Chrome and Safari users to visit a particular URL is no big deal really.

Maybe its not that simple, there are few things to consider here. When you send the first request to a particular page and the response does not contain the 'Access-Control-Allow-Origin' header with a suitable value then the browser refuses to send more requests to the same URL. This however can be easily bypassed by making every request unique by adding a dummy query-string parameter with changing values. The number of requests/minute is also a variable. The browser sends a certain number of requests and when it receives the responses for those it sends in the next set of requests and so on. So as the server slows down the browser's requests/minute rating would also slow down. The figure 10,000 requests/minute was clocked against a server located in the internal network, against a target in the Internet it would realistically be between 3000-4000 requests/minute. If the attacker is planning to target an internal server by getting the employees of that company to visit this malicious URL then the 10,000 requests/minute rating would apply.

I am not going to release any PoC as this might probably be a bad time to do that but it shouldn't be very difficult to put together something for testing once you understand how it works. It should be relatively easy to block this attack at the WAF since all Cross Origin Requests contain the 'Origin' header, that way you can differentiate between legitimate and malicious requests.

Saturday, December 11, 2010

Port Scanning with HTML5 and JS-Recon

This was one of the newer topics that I covered at BlackHat Abu Dhabi. HTML5 has two APIs for making cross domain calls - Cross Origin Requests and WebSockets. By using them JavaScript can make connections to any IP and to any port(apart from blocked ports), making them ideal candidates for port scanning.

Both the APIs have the 'readyState' property that indicates the status of the connection at a given time. The time duration for which a specific readyState value lasts has been found to vary based on the status of the target port to which the connection is being made. This means that by observing this difference in behavior we can determine if the port being connected to is open, closed or filtered. For Cross Origin Requests it is the duration of readyState 1 and for WebSockets it is readyState 0.

I tried to do some calibration of the time duration for the different port states and the data is below. These numbers only hold good when the target is in the internal network. If you are scanning a target on the internet then the network latency should be taken in to account.

Since this is not a socket-level but an application-level scan the success also depends on the nature of the application running on the target ports. When a request is sent to certain type of applications they read the request and remain silent keeping the socket open, probably expecting more input or input in the format they expect. If the target is running such a application then its status cannot be determined.

Since even closed ports can be identified we can extend this technique to perform network scanning as well as internal IP detection. I have written a tool called JS-Recon which can perform these. More details on the how JS-Recon works is here. These techniques only work when run from Windows machines, on *nix systems it is not possible to determine closed ports and the timing figures are quite different.

Tuesday, December 7, 2010

RSnake, Web Security and a few beers

BlackHat Abu Dhabi 2010 is special to me for many reasons, chief amongst them is that I got to meet one of my most favorite hackers - RSnake. When I started taking my baby steps in web security, like most people at that time, I started off by reading the excellent content available on OWASP combined with some heavy use of Google. Around this time I think it was Manish that introduced me to ha.ckers.org, am so glad that he did, it almost immediately became an addiction. Back then I only understood parts of what was written on ha.ckers.org, sometimes an entire post would be beyond my grasp but I still enjoyed reading them. It was not just a blog it was an event, an event where RSnake and his loyal band of commenters had a gala time. There are many instances where I had read a few Wikipedia articles and docs to understand a topic just so that I could know what RSnake and the commenters were laughing about.

I did not have to follow the RSS feeds of a few dozen blogs instead I only checked ha.ckers.org on a regular basis. If there was some interesting news in the web security world then it would be talked about at ha.ckers.org along with RSnake's opinion on how significant it is and how it impacts things, served with a pinch of humor. This was in addition to RSnakes's own bag of tricks which always had something clever. ha.ckers.org was an excellent learning medium and has probably helped and inspired countless folks like me across the world. Its is extremely hard to discuss an advanced topic without making a novice reader feel alienated and bored. Also it is equally hard to discuss a technically simple yet important topic without making the smart ones cringe. Somehow ha.ckers.org managed to do both very well, a feat that is very hard to match.

Coming from a part of the world where you almost never get to meet most of the famous hackers in person, in our heads RSnake usually has a larger than life image, he is more like a WebAppSec folk hero. So meeting him personally was really special. As a person he is very friendly, chilled out and did not seem to mind the fact that I am relatively a n00b :D. We spoke for quite a long time, heard a lot of interesting stories related to ha.ckers.org, his book and more. Though he didn't seem to like my choice of beer, meeting him has only increased my respect for him. He is one of the key figures who has shaped the web security industry and an inspiration for many.

This is an excerpt from a recent interview of his:
..if you love security, don't let the people at the top of the security industry dictate the terms by which you do your research, disclose your vulnerabilities, or do your job.You have a ton of potential, and life is too short.My Father used to tell me that if you love what you're doing you'll never work another day in your life.To paraphrase him - if you aren't having fun in security, you're doing something wrong.Put a smile on your face, and go do what makes you happy!
This probably says more about him than I can in a few dozen posts. As he shuts down ha.ckers.org to go on a different journey I would like to wish him success on behalf of all his followers from India. Good luck RSnake!