Showing posts with label DDoS. Show all posts
Showing posts with label DDoS. Show all posts

Wednesday, August 16, 2017

New Research on "Pulse Wave" DDoS Attacks



New findings from Imperva Incapsula researchers published today Attackers Use DDoS Pulses to Pin Down Multiple Targets details the emergence of a new assault pattern, which they’ve named Pulse Wave.   

According to lead researcher Igal Zeifman, “Pulse Wave DDoS represents a new attack methodology, made up of a series of short-lived pulses occurring in clockwork-like succession, which accounts for some of the most ferocious DDoS attacks we mitigated in the second quarter of 2017. In the most extreme cases, they lasted for days at a time and scaled as high as 350 Gbps.”

The size of these attacks, and the amount of skill they exhibit, are likely the handiwork of skilled bad actors who have become practiced in portioning their attack resources to launch simultaneous assaults — meaning the intervals between each pulse are being used to attack a secondary target.

This new approach shows that some offenders have grown to understand that it is not necessary to hit a target continuously to take it offline; rather, repeated short bursts are enough to disrupt routers and servers, producing the same effect. By the time the systems have recovered from the first burst, or pulse, the hackers hit them again. In this way, they can double their resource utilization and pin down several targets.
 


The existence of such capabilities spells bad news for everyone, as they enable bad actors to greatly increase their attack output. The pulse-like nature of these attacks, however, is especially harmful for appliance-first mitigation solutions, since it can cut down the communication between their two components, preventing effective failover from the appliance to the cloud. Specifically, the attacks have the capacity to delay the time it takes for the cloud component of the mitigation solution to kick in. This increases the likelihood of the target going down and being forced to initiate a prolonged recovery process.  Moreover, the pulse wave assaults can prevent transition of data collected in the early attack stages from the appliance and into the cloud to further harm its responsiveness.
As the research points out, while Pulse Wave attacks constitute a new attack method and have a distinct purpose, they haven’t emerged in a vacuum. Instead, they’re a product of the times and should be viewed in the context of a broader shift toward shorter-duration DDoS attacks. Multiple industry reports—including the Imperva Incapsula quarterly DDoS Threat Landscape report— point to an increased number of short-lived DDoS events over the past year. As a result, the majority of all DDoS attacks today —both at the network and application layers— consistently last less than one hour. Moreover, the percentage of such short-burst attacks is growing each quarter. 

“For a commercial organization, every such instance translates into tens of thousands of dollars in direct and indirect damages. For professional offenders—already inclined to split up their attack resources for optimized utilization—this serves as another reason for them to launch Pulse Wave DDoS assaults. Consequently, we expect to continue encountering such assaults. We also forecast them to grow larger and become more persistent, fuelled by botnet resource evolution and the previously described macro trends we’ve observed in the DDoS landscape,” Zeifman added.

The full Research Paper ”Attackers Use DDoS Pulses to Pin Down Multiple Targets, Send Shock Waves” presents a detailed dive into the nature of pulse wave attacks and the threat that they pose and their place in the DDoS threat ecosystem.

Monday, January 9, 2017

Three Critical 2017 Predictions for Global Cyber Security



Three Critical 2017 Predictions for Global Cyber Security
Leo Taddeo, Chief Security Officer, Cryptzone

The US will take a more aggressive position on international cyber security that will lead to cyber escalation between Nation States.

Typical to any transition of Presidential power, we’re in a honeymoon period between Nation States. Make no mistake, however: the first thing our adversaries are doing is figuring out where to “pick a fight” to see what they can get away with. 

For example, Russia has a track record of positive initial meetings with President-elects, while they test how much they can extract from them. But soon enough, Russia’s concerns will be intractable and Trump will be forced to face reality: Russia will not change its behavior.

We are already seeing Russia test Trump. Just this week, Putin said about Russia’s military, “We can say with certainty: We are stronger now than any potential aggressor. Anyone!” These kinds of comments will not sit well with an aggressive President-elect, and a Cabinet full of former US military.

With relations strained between Trump and the Obama Administration, negotiations between the US and Russia have already been sent into a tailspin. All of this will lead to cyber escalation.

While Trump has been playing nice – some say too nicely – with Russia, he’ll eventually overact and take proactive measures, which is more his style.

Another “Russia intractable” is the Ukraine. Because the US having influence in Ukraine is threatening to Russia, and the US won’t leave Ukraine’s fate in Russia’s hands, both countries will use cyber tools to try and get what they want: information.

China is another example. Cyber relations with China are already unstable with the incoming US leadership. Trump is taking China head on, with regular comments on its currency devaluation, abuse of trade policies and the volatile Taiwan relationship.

In response to a new, perhaps more hostile Trade Secretary,  China has indicated that “China like every other country is closely watching the policy direction the US is going to take.” 

In September 2015, Obama agreed with China that neither side would engage in cyber espionage in business. It was a gentleman’s agreement, based on goodwill, that isn’t binding or enforceable. From detectable cyber activity, there seems to have been a decrease in cyber espionage, supporting the notion that both sides have been honoring the agreement. If that goodwill doesn’t exist with China, and it seems that it does not in the President-elect’s approach, all bets are likely off.

The world is not equipped to handle sophisticated, multi-site cyber attacks, especially against financial institutions. Stolen money will be reinvested in “hacker R&D” creating future chaos. 

Countries and corporations are not prepared to deal with advanced cyber attacks. In the February 2016 Bank of Bangladesh hack against the SWIFT system, criminals stole $81M—and most of it is still unrecovered. Hackers are already re-investing these funds to develop techniques to target lesser protected institutions, which isn’t good news for network defenders.

The IoT is becoming even more commonplace and the lack of set standards and regulations leave us with more and more unsecured devices, widening the playing field of opportunity for hackers.

DDoS is becoming sexy again because we’re entering a different era in terms of volume, in part due to the number of IoT devices now online. We should anticipate an acceleration in DDoS attacks because some of these devices simply can't be fixed or properly secured. 

The October 2016 DYN DDoS attack is a good example of the above two trends. And the IoT botnet (Mirai) used in this attack shows signs of evolving as its source code was released publicly.

Companies will soon make the official transition to cloud, as they’ll stop viewing it as a risk and more of a sanctuary. They’ll also start establishing a “TSA-type security pre-check in line” to services for approved clients that will isolate channels for customers (i.e., they won’t be public facing) in order to avoid the Internet cesspool. 

Context is the next evolution of identity and people will, finally, stop caring about giving up privacy in order to prevent attacks. 

We’ll see identity finally move beyond a username and password: things like what device you’re on, why or what’s the context, enterprise vs. mobile origination, etc. that are seamless (invisible to the user) will take precedent and they’ll be embedded in use, and travel wherever users do.

The resource (device, company, network, app, etc.) will care about who you are in the move to cloud and BYOD environments. As part of this, users will give up privacy to access the resource.

This trade-off is fair. You must provide enough proof of who you are when asking for access to a valuable, shared resource. Users already sign end user license agreements, which most don’t read, and scroll as fast as possible to click accept, granting necessary access. The vast majority of the population views this as a fair and acceptable trade.

About the Author
Leo Taddeo, former Special Agent in Charge of the FBI's NY cybercrime office, is now member of the Citizens Crime Commission of NYC and CSO at Cryptzone.

Thursday, October 27, 2016

New Stats on Dyn DDoS Attack Size



Imperva Releases More Information on the Dyn Attack

Ofer Gayer, product manager at Imperva for the Incapsula product line, explains:

“There is still quite a bit of speculation swirling on the size of the DDoS attack on Dyn last Friday. We know there were 100,000 Mirai botnet nodes – which is not especially large in our experience. So, in our estimation, there are two likely causes. The attack may have been a high-volume attack – over 500 million packets per second – that overwhelmed the Dyn infrastructure. Or, the attack may have been relatively small – 50-100 million packets per second – and the attack itself was “amplified” by what is known as a retry storm from their millions of legitimate users, making the job of differentiating between good and bad traffic very hard.”
 
Additional Information:

Q. Is a 100,000-node botnet big?
A. Not really.  Example of a 180,000-node botnet mitigated  https://www.incapsula.com/blog/headless-browser-ddos.html

Q. Are DNS services especially vulnerable?
A. They do suffer from being open systems:

"Effective DDoS mitigation is synonymous with accurate traffic filtering. For that reason DNS amplification attacks are actually easier to deflate as all uninitiated DNS responses are highly suspect and could be filtered on-edge, without any impact on the regular traffic flow. For example, one could categorically drop all unexpected DNS responses to port 53.

However, this isn’t the case for seemingly legitimate DNS flood queries, which cannot be dismissed before they are individually processed at the server level.

With on-edge filtering bypassed, and the path to the server CPU cores laid wide open, DNS floods have the potential to bring down even the most resilient of networks. "


Q. How can companies prevent attacks on their DNS infrastructure?

Q. Is Mirai that sophisticated?

Q. Has the Incapsula network been hit with Mirai?

Q. What’s a big DDoS attack measured in million packets per second (Mpps)