Latest Stories

Stay up-to-date with everything at Approach

Blog article

Rise of DDoS Amplification Attacks

Publication date

04.04.2018

Discover how easily DDoS amplification attacks can cripple networks, why they’re harder to stop than you think, and how you can protect your infrastructure before it’s too late!

The DDoS Context

“A denial-of-service attack (DoS attack) is a cyber-attack in which the perpetrator seeks to make a machine or network resource unavailable to its intended users. It temporarily or indefinitely disrupts services of a host connected to the internet’ – Wikipedia 

In the case of distributed denial-of-service attack (DDoS attack), the load of the attack is originating from many different sources. It generally uses a botnet and/or bad configured services (as we will see further). Unlike DoS, that type of attack makes difficult tracking the hacker and involves enormous bandwidth consumption of the victim.

There are many reasons for this type of attack, including the following:

Ideological: some kind of objection to a message conveyed by a company or a government (e.g.: Anonymous)

Smokescreen: used to hide a more complex attack, involving generally impersonation or data exfiltration

Challenge: to show a capability of a hacker to overcome or to size the security of a system

What’s more frightening is the ease of making a DDoS for cheap, even for someone with limited skills, because the artillery used is widely available for free or almost free.

What’s an amplification attack in a DDoS context? 

An amplification attack is a volumetric attack trying to exhaust the bandwidth of the victim through “reflection”. An attacker will “ask” with a short message an information, in UDP, to a vulnerable service. The attacker will replace the source IP of the message with the IP of the victim, to convince the vulnerable service to “respond” directly to the victim. When the service responds with a message larger than the source message, it’s called amplification. The three-way handshake of TCP doesn’t allow that behaviour, as the source address cannot be spoofed – Wikipedia.

In this figure, we have described the specific case of a hacker sending directly a spoofed packet to several vulnerable services. In large scale attack, the hacker will use a botnet to send the initial packet from a big amount of sources, which will exponentially increase the effect of the attack.

The potential effect of an amplification attack can be measured by the Bandwidth Amplification Factor (BAF), which can be calculated as the number of UDP payload bytes that an amplifier sends to answer a request, compared to the number of UDP payload bytes of the request (see below a list of known protocols and their associated BAFs.

Protocol Bandwidth Amplification Factor
DNS 28 to 54
NTP 556.9
SNMPv2 6.3
NetBIOS 3.8
SSDP 30.8
CharGEN 358.8
QOTD 140.3
BitTorrent 3.8
Kad 16.3
Quake Network Protocol 63.9
Steam Protocol 5.5
Multicast DNS (mDNS) 2 to 10
RIPv1 131.24
Portmap (RPCbind) 7 to 28
LDAP 46 to 55
CLDAP 56 to 70
TFTP 60
Memcached 10,000 to 51,000

Prevention and Response

Here are some points that can help you in mitigating quickly an amplification attack:

– An up to date network/service diagram.

– A tested incident response plan with roles and contacts well defined (ideally also printed on paper, in case of unavailability of the whole network).

– A proper monitoring system that will alert you at the start of a DDoS.

– A firewall that block all traffic that is not explicitly permitted (incoming and outgoing).

– Use a properly configured load balancer and make sure it does not become a single point of failure.

Assess regularly the security of your infrastructure and of your firewall rules.

– Ask your provider (ISP) for some temporary filtering (traffic scrubbing or traffic shaping).

– Take note of every action during an incident response (timestamp, actions, results, …).

Escalate to a third party if no other solutions.

And … do not participate in a DDoS.   Prevent open or misconfigured UDP services to avoid participating in a future DDoS.

Our statistics for Belgium

The services that can be used for an Amplification attack are widely available all around the world. To give you an idea, we have compiled statistics on some vulnerable UDP service in Belgium, coming from Threat Intelligence and passive scans at Belgian level, as of mid-March 2018, that can be used for Amplification Attack (ordered by amplification factor):

Protocol Bandwidth Amplification Factor Vulnerability in Belgium
Memcached 10,000 to 51,000 55
NTP 556.9 10809
CharGEN 358.8 26
QOTD 140.3 49
DNS 28 to 54 2743
SSDP 30.8 2757
MS-SQL 25 537
Portmap (RPCbind) 7 to 28 6025
SNMPv2 6.3 5993
Multicast DNS (mDNS) 2 to 10 6093
NetBIOS 3.8 2683

Conclusion

– DDoS Amplification Attacks are not a matter of skills for a hacker, since there is a plethora of tools, more or less freely accessible.

– DDoS Amplification Attacks are difficult to mitigate since the IP of the hacker is hidden behind the reflection.

– Huge quantity of vulnerable services that can be used for the DDoS are available all around the world. At least beware to not participate yourself.

– You can the “Prevention and Response” section as inspiration for your own defence.

 

OTHER STORIES

Cybercriminals keep evolving—uncover the latest malware delivery tricks, evasion tactics, and real-world attack chains to stay ahead in cyber security.
DNS over HTTPS (DoH) boosts privacy but opens new security risks—learn how cybercriminals exploit it and how enterprises can stay protected.
SSH security is at risk with the Terrapin attack—learn how it exploits vulnerabilities, weakens encryption, and what steps you need to take to stay protected.

Contact us to learn more about our services and solutions

Our team will help you start your journey towards cyber serenity

Do you prefer to send us an email?