Hackers Battle For 3 Million Strong Mirai Botnet

By Brian Karas, Published Nov 28, 2016, 10:14am EST

Mirai-infected devices have become so large and so prevalent that multiple hackers are now fighting each other to control these devices.

This war has both made Mirai wider-spread, but less powerful for any given attack, and may lead hackers to search for new vulnerabilities in cameras and recorders to grow their botnet army.

In this report we look at the latest status of Mirai, and how it is expected to evolve into 2017.

3 ******* ******

***** *** **+ ******** ******* (most ****** *** ** different *******), **** **** * million ******** ******* **********, and ~***,***+ ************* ** attacks ** *** ***** day, ********* ** *** *********** **** [link ** ****** *********], run ** * ****** researcher ** *** **, and ****** *** ****** of *****-***** ******* ********* daily.

***** *** ** ********** infected ******* ***** **** distribution ********, ********* *** US *** ******:

Popularity ****** ********** *****

** **** ***** ******* have ****** ** **** the *********** ** ***** botnets *** ******** **** other *** ******* ** vulnerable *******. **** ******** ************* ********, ** *** inevitable **** **** ****** controllers ***** *****, ***** has ********. **** *** less *********** *** *** way **** ***** ****** the ********* ********* *** any ******** ******. * recent***** ******* ** ***** ****** **** ****** ********* are ******** ********* **** other ** *** *** take ******** ********* ******* and **** ******* ** more *******:

*** ******** ********* ** use *** ***** ******’* resources **** **** ******** to turning ***** **** ******* ** each ***** ** *** to **** ** ******.

*********** **** **** *********** this ******* ******* ** botnet *********, *** *** ** is ********* ***

** ***** *** ***** hackers ** ****** *****-***** DDOS ******* ************ — such ** ***** **’** seen ******* *****, ***, and *** — **** will **** ** ********** competitors *** *** ***** to ******* *** **** type ** *******. ****** not **********, ********** **** competition ** * ************ challenge *** ** ***, the *** ****** ********* appears ** ** ********* with *** **** *****-** controllers *** *** ****** new ********** *******.

98 ******* ** *********

**** ******** ****** ***** the ******** *** ***** exploitable *******, **** ******* army ** ******** ***** that ********** ******* **** be ******** ****** ******* of ***** ******* **. One ********** ********* * camera, ************ ** ********, on ****** *** *** purposes ** ******** *** Mirai ** *********** *** it****** ******** ** ** seconds******** ******* **** ** commonly ****** **** *** company ****** * ******, *** ********** ******* ******* ***** blaming it *** *** *************** that *** ** *****.

Mirai ***** ** ******

*** ** *** *** Mirai *****, *********** * fresh **** ** *** exploit **** **** **** newly ******** ******, ** is **** *** ********* to ****** *** ****** code **** *** ******** or *************. **** ******* believe ****, ******* **** the *** *** * relatively ******* ****** ** resources, **** ***** *** attack **** ** ****** to ** **** ** infect ******* ********** ******. This ***** ** **** by ********* *** ********/******** list ** ******* **** than **** ******* *********, but **** **** ****** passwords. *** **** ***** also ****** ** **** more ***** *** **** telnet **********, ** ****** devices ******* ****** ** non-standard *****.

Outlook *** ****

** *** *** *** another ******-******** ****** **** Mirai, *** **** ****** see ********* *******, ** lower-scale ******* **** *** carried *** *** ****** periods ** ****. *** cost ** ******* *** a ****** *** ****** may drop, *** ** *********** from *** ******** ********* competing **** **** ***** for **** ********, ******* to **** ******* ** lower-profile *******. ** ********* manage ** **** * new ***** ** ******* to ****** ** ***** return ***** ** ******** record-breaking ******, ********** ** a **** ***** ******** than ******.

Comments (13)

This is a really interesting article, great job. My question in follow up to this article, is there even much that can be done to prevent this type of stuff from happening? The cat has been out of the bag for so long, and there so many criminal and militaristic (irony?) opportunities for these networks, that I see no fix to these types of bot/zombie nets. You can try to avoid buying products that can be cracked, but theres new vulnerabilities every week, its a losing proposition.

From a network security standpoint there are Firewalls from PaloAlto/Cisco/Fortinet/Baracuda/etc that can detect and deny DDOS traffic attacks, prevent your botnet infected devices from communicating out, but those tools on high enough throughput hardware can run in the multithousands of dollars price range.

As you mention, strong passwords could be a huge fix to a large chuck of these botnets, but we all know that's a lost effort. Unless you can overhaul the entire alphanumeric password string paradigm (you cant).

Agree
Disagree
Informative: 1
Unhelpful
Funny

My question in follow up to this article, is there even much that can be done to prevent this type of stuff from happening?

I think there are a few things that are easy and help reduce the problem:

  • Manufacturers can disable telnet and SSH by default (to be fair, it seems like this is much more common now).
  • User authentication methods could be updated to not allow remote login if a secure password has not been set (this would allow local login for setup/configuration, but deny remote access until a better password was used).
  • Installers can help by not forwarding excess ports, and also putting services on non-standard ports. Using non-standard ports makes it slightly harder to scan for services to find devices that can be further probed.
  • Firmware updates can be done as part of the installation process to ensure products are running latest code, which in most cases should have reduced vulnerabilities, though it is always possible that new vulnerabilities have been introduced, but this should be rare.
  • Installers can also select and recommend products based on their ability to be secured, and the vendor having a good track record of responding to and fixing reported issues quickly.

Manufacturers could make a huge leap in reducing device exploitability if they were willing to make it a priority, and willing to invest proper effort into building more robust devices, by adding secure boot capabilities.

Some SoC suppliers, such as Ambarella, have been adding secure boot functionality into their chips, but it does not appear that security manufacturers are doing anything with it.

In short, secure boot can be utilized to ensure that even if an attacker could gain access to a root shell, they could not load/run unauthorized code.

Secure boot would make it several orders of magnitude more difficult to create botnet's, as the hackers could not just download software to make the camera/recorder do whatever they want. This is roughly similar in concept to how an iPhone cannot run software that has not been officially vetted by Apple, meaning that manufacturers like Axis would not have to give up the ability to run 3rd party apps on their cameras.

Like most things, creating software that is more secure and goes through additional authentication processes would take more effort on the manufacturers or developers part, but the result would be significantly increased security, and trustability, of recorders and cameras. But, the manufacturers are only likely to add this in if there is financial benefit to doing so, either by customers requesting it, or the support headaches of hacked devices becoming significant.

Agree: 1
Disagree
Informative: 2
Unhelpful
Funny

You mention installers a few times in your response, but my question is arent a lot of the devices that are part of these botnets actually consumer facing devices?

Also from that same perspective, how many of these limitations, extra steps would be accepted by consumers, and followed up on, and not just generate complaints that their Web connected video chat camera wouldn't talk to the cloud service?

From a manufacturers standpoint, your last comment says everything I think I need to know, with a 3rd possible option for manufacturers adding it; regulations (gasp!)

Agree
Disagree
Informative
Unhelpful
Funny

You mention installers a few times in your response, but my question is arent a lot of the devices that are part of these botnets actually consumer facing devices?

Yes, it is very likely a lot of them are consumer/DIY installs, I just listed the installer suggestions because it is relevant here. I had also considered listing some suggestions for end-users, but I think the reality is that very few end-users are aware enough to do anything in the first place, and those that are aware already know common best practices.

Agree
Disagree
Informative
Unhelpful
Funny

User authentication methods could be updated to not allow remote login if a secure password has not been set (this would allow local login for setup/configuration, but deny remote access until a better password was used).

Yes, couldn't agree more: Simple Solution To Default Password Conundrum...

Agree: 1
Disagree
Informative
Unhelpful
Funny

The problem that will continue to exist is that there are hundreds of thousands of these cameras that are already out there and that will not be updated due to ignorance of the problem, not wanting to revisit old installations, or just not caring. For most products sold in the last year or so, the vulnerability was addressed by closing the telnet ports to deter infection of the devices. This botnet will continue to function due to the old cameras that will not be updated, not due to new cameras being purchased today or even within the past several months.

Agree: 2
Disagree
Informative
Unhelpful
Funny
Any other possible solutions? Say camera chip built in software to deny (control) MOST Wan access? OR A firewall configuration to deny selected local (camera) LAN IP's WAN access? NO, I don't know how to do that. Just seems an honest question. OR is that already available. If so, where does one find it?
Agree
Disagree
Informative
Unhelpful
Funny

What you are describing is already available with either standard firewall configs, or in some cases cameras allow for "whitelist" IP access, where you can restrict which remote machines can establish a network connection to the camera. The downside to this is that it makes access via mobile apps much more challenging (unless the manufacturer, or some other entity, provides a cloud access/proxy service). I can also be difficult if the IP address of the remote access site changes for some reason, which is not uncommon for residential cable/DSL accounts.

Agree
Disagree
Informative
Unhelpful
Funny

Thanks for the explain.

My question was based on the bot having already gaining access. An issue surely. But once in...

On the simple side (for the average homeowner with a single camera or two) A firewall CAN deny the already in place bot from gaining WAN access. Thus thwarting it's mission? Next would be to have that information provided to the volume of such working cameras and possibly reduce a large part of the problem identified in the post.

Not a total solution granted but a reasonable attempt to minimize the problem until something more permanent comes along and one IPVM could champion from it's access to the world.

Agree
Disagree
Informative
Unhelpful
Funny

All reports that I've seen about the nature of the infection says that once the botnet agent infects a device, it removes the downloaded code and only stays resident in memory. Power cycling the device would remove the botnet agent, but unless measures are taken to protect it before powering back on it will most likely be infected again within minutes.

Agree: 1
Disagree
Informative
Unhelpful
Funny

Very informative which I was not aware of.

Agree
Disagree
Informative
Unhelpful
Funny

The company I work for has an very aggressive ACC to reduce the risk of hacking. The problem will always be human error and the insider threat.

Agree
Disagree
Informative
Unhelpful
Funny

IPVM Image

Agree
Disagree
Informative: 1
Unhelpful
Funny
Read this IPVM report for free.

This article is part of IPVM's 6,910 reports, 923 tests and is only available to members. To get a one-time preview of our work, enter your work email to access the full article.

Already a member? Login here | Join now
Loading Related Reports