Nice article. If we kee pushing the industry and best practices, these types of hacks will surely be minimized. Manufacturers, of course, play a large role here, but integrators have to do their part too. VPNs aren't always an option. Sometimes a device needs to be public facing. Restricting the access to that public facing device and monitoring traffic to it is essential.
Interview With Researcher Who Cracked Security Of 70+ DVR Brands
[link no longer available]The researcher who found an exploit in 70+ brands of TVT OEM DVRs, which he suspected to be the entry point hackers used to gain access to several retailers, elaborated on these hacks to us.
We got in touch with Rotem Kerner [link no longer available], previously a security researcher at RSA, and now co-founder of security research company Cybewrite to discuss the exploitability of embedded security devices.
Old **** ******* *** * ***** ******
****** **** **** ** ***** ************* of ******** ******* **** ********* "****** on *** **** **** *** * lot ** ****** *** ****** ** hack ****". ******** ** ***** ******** devices **** ***** ******* ** ***** *******, ******** cameras *** ********* *** *** **** nearly ** **** ***** ********, *** accessible ** ********, ****** ************* **** concerned ***** ***** ********.
Now ****** ********
*** ******* ******* *** *********** *******, security ******* **** ***-***** *******, ** ***** up ***** *** **** *** *****. With *** ******* **** ** ** camera ******* *** ********* ********** ** Internet-connected ******* **** **** ****** *********, *** easy, *******.
Worse ** ***** ******* ** ********
** ***** ******* ***** **** ** the **** **** ****** *** ******** http ******* *** ************* *** **********. **** can **** *************** ***** ** *** device *********** ********** ** ***** ******/******** using *** **** ******** ******** *********. Another ****** ************* ****** ****** ******* ******* ** ******** ****** ****** all ******* ****** ********.
How *************** *** *****
***** ***** ****** ** *********** ******** images, *** ***** ***** **** ******* ** ********* **** *** **** ** the *** ***** ** *** ********. **** runs ******** ** *** ************ ********** ** ******** * device ************ *** **** ***** *** trying ******** ******* **** *** ********** ** *** ************.
********* ** **** ** *** *** hacks, ************* *** ******** ***** ********, also ********* * ********* **** ** brands/logos. **** ******* *** ** *********** determine **** *** ******* ** ******* could ** ******* ** ********* ** deployed ******* **** **** ** ******.
Two ****** ******* ** ******* ********
** ******* *** ** ********* ** a ****** ** ******* **** ** shell ****** ** * ***** ******* and **** *********** **** ******** **** the **** **** * ****** ******. This ****** ******** *** ****** ** be **** ** *** ******* ****** to *** **** (****** *******, ** over *** ********). ** *** *** benefit ** ***** **** ** ** done ******* *** **** ****** ** do ********, *** ********** *** ******* to ********** ******** ** ******* ******** or ***** **** *** ********** ********.
******* *** ** *** ** ******* onto * **** ** ** ********* new ******** *** *** ****** **** has *** ******* ***** **, *** then ********* **** ******** **** *** unit. **** *** **** ** **** remotely, ** *** ****** *** *** and *** *** **** ** ******* the ******** **********. *** *** ** convince * **** ** ******* *********** firmware ** ** **** *** ******** available *** ********, *** **** **** an ***** **** ******* ** **** from *** ************* ******* ********** ********** a ******** ******.
Securing **** *******
** ** ********** ** ***** ****** a ****** ********* ** *** ********, but *** *** ****** **** ******* of ****** ** ********* ***** **********:
- ****** ******* *********, *** ********* ** the ****** ****** **. ****** ******** "admin" ** "*****" **** **** ****** ********** ****** ********.
- **** ******** ******** ***** **** ************* websites, ***** **** **** *** ********* files *** *** ***********, ** **** reduce *** ******* ** ******* ***** firmware.
- **** ******** ******* ** ******* *** exploits **** ***** ****, ******** **** the ************ ** ******* ** **** security *******.
- ***** *** ** *********/******** **** *** access **** *********, ** ****** *** close ****** ******** ****** *** ******* via ***.
** *** ****** **** ******** ************* ************, ********** ******** ****** ** ******* can ****** **** ** *******.
What ************* *** **
******** ****** *** *** ******** **** the ******** ********** *** *** **** common ******* ** *******. ******* ** ***** access ** ****** ****** ** ********** use *****, *** ** ***** **** by ******* ** **** ****** ** the ******.
************ ******** ********* ******** ****** ************ removes * ***** ****** ******. **** ******* ** not ****** *** ********* ** ******** files ****** ********** ****, ******** ********* to ****** ***** ******** ****** ***** *** turn ***** ******* **** *********** ******* access *****.
********** *** ******** ***** *** **** make ** **** ********* *** ******* to ********* **** *** ***** *** code *** *********** **********.
Like it!
Confused about this though:
This allowed him to immediately determine that the exploit he created could be applied to thousands of deployed devices from over 70 brands...
An exploit can be installed on a device by gaining root or shell access to a linux console and then downloading code directly into the unit from a remote server.
Isn't the 'exploit' the ability to 'gain root' in the first place?
Sounds to me like you are saying, "Cars can be stolen by using the key to start the engine."
Isn't the 'exploit' the ability to 'gain root' in the first place?
In some sense, yes. In this case I was referring more to the exploit being the unintended code that was uploaded to the device and then executed by the device. But just the act of gaining root/shell access is also a form of exploit.
Isn't the 'exploit' the ability to 'gain root' in the first place?
Semantics, but as I see it, the ability to gain unauthorized "root" is the hack. The "exploit" is the change/action you make/take (usually nefarious) as a result of gaining unauthorized access-often installing malware.
The "exploit" is the change/action you make/take (usually nefarious) as a result of gaining unauthorized access-often installing malware.
Disagree. Though that makes sense in the common understanding of the word 'exploit', I don't think that is the usual meaning in cyber-security.
For instance if I use a brute force password program to gain root to a system and then I burn a DVD of all the credit card numbers, the exploit is brute force password attack, not copying files.
This makes sense, since otherwise the exploit could be any number of things, that could be done once you are root.
In this particular case, the result of "exploit he created" is nothing more than the root access itself. This is the usual meaning.
This was one of the two researchers whom I originally thought may have uncovered the Axis vulnerability. I ruled him out on that one and am happy to see this kind of coverage on IPVM.
Is there a way to know if a device has already been compromised? Or will defaulting and then upgrading to latest firmware eliminate any rogue hack placed on the device?
This is an excellent question.
The generally accepted answer is that there is no way to know for sure if a device has been compromised, especially when it comes to things like cameras where tools like chkrootkit are not yet adapted.
Defaulting a camera will only erase user-configured settings in most cases, it won't wipe out any rogue software that has been added.
The good part is that for most devices the firmware contains a complete image, operating system and all additional files/software needed to run the camera or recorder. Upgrading firmware should do a complete wipe and overwrite any malware. However, just like configuration settings can be saved across a firmware upgrade, so could installed malware if the hackers were particularly sophisticated and understood the internals of the device fully.
I do not think we are at the stage where malware targeted to cameras and recorders is at the point where it can survive across a reset/default and a firmware upgrade, but it is technically possible for this to occur. Hopefully we will see more advanced firmware and code checking implemented by manufacturers before the hackers get ahead of them in this regard.
I've got backdoors for DVRs that persist through a firmware upgrade. Most firmware updates don't touch the bootloader.
Considering the brands hacked, maybe it is time to quit selling the low end crap and start designing systems and networks with actual security in mind. You know, that thing we are supposed to do in the beginning.
Salespeople need to quit being price conscious and start selling the service expected of our industry. If end users are so price sensitive, it is time to educate them better to make better informed decisions about their security.
While I agree in general with your sentiments, I would like to know which brands are exempt from exploits, in your opinion?
So when you refer to 'low end crap', do you mean that higher price = cyber secure?
It isn't the brands, per se, but how the systems are designed. Working with the customers' IT people can limit exploits, dedicated security networks. Or for the safest route, a system that is not connected to the internet, except in cases of service and maintenance.
However, using shells that anyone can write a GUI for can lead to exploits from outside sources, due to not maintaining a standard of care for securing the code.
Staying with companies whose software is vetted to limit potential exploits is part of the due diligence integrators must look at when selecting a VMS.
Can you help me identify "companies who's software is vetted to limit potential exploits" vs ones who don't?
Over the years, I and my co-workers have done that due diligence. I have systems I like, and systems I do not. But sharing that information is something I keep in house and treat as a intellectual property,.
Whatever dude. Carry on with your systems, and I will do the same with mine.
Thank you Jon, you hit the nail on the head for sure! Comments like those are what leads to all of the "mis-information" we deal with.
There is a new botnet being reported, called "Tsunami", that looks to be targeting the TVT devices. So far ~227,000 vulnerable devices have been identified: