Of all the Q4 2015 events in the world of DDoS attacks and the tools used to launch them, we picked out those that, in our opinion, best illustrate the main trends behind the evolution of these threats.
- Emergence of new vectors for conducting reflection DDoS attacks;
- Increase in number of botnets composed of vulnerable IoT devices;
- Application-level attacks – the workhorse behind DDoS attack scenarios.
Attacks using compromised web applications powered by WordPress
Web resources powered by the WordPress content management system (CMS) are popular with cybercriminals who carry out DDoS attacks. This is because WordPress supports the pingback function that notifies the author of a post published on a WordPress site when someone else links to that post on another site running the same CMS. When the post containing the link to the other web resource is published on a site with the enabled pingback function, a special XML-RPC request is sent to the site where the link leads and that resource receives and processes it. During processing, the recipient site may call the source of the request to check for the presence of the content.
This technology allows a web resource (victim) to be attacked: a bot sends a specially formed pingback request specifying the address of the victim resource as the sender to a WordPress site with the pingback function enabled. The WordPress site processes the request from the bot and sends the reply to the victim’s address. By sending pingback requests with the victim’s address to lots of WordPress resources with pingback enabled, the attackers create a substantial load on the victim resource. This is why web resources running WordPress with the pingback function enabled are of interest to cybercriminals.
In Q4 2015, resources in 69 countries were targeted by DDoS attacks #KLReportTweet
The power of one such DDoS attack registered by Kaspersky Lab experts amounted to 400 Mbit/sec and lasted 10 hours. The attackers used a compromised web application running WordPress as well as an encrypted connection to complicate traffic filtering.
In October 2015, experts registered a huge number of HTTP requests (up to 20,000 requests per second) coming from CCTV cameras. The researchers identified about 900 cameras around the world that formed a botnet used for DDoS attacks. The experts warn that in the near future new botnets utilizing vulnerable IoT devices will appear.
Three new vectors for carrying out reflection DDoS attacks
Reflection DDoS attacks exploit weaknesses in a third party’s configuration to amplify an attack. In Q4, three new amplification channels were discovered. The attackers send traffic to the targeted sites via NetBIOS name servers, domain controller PRC services connected via a dynamic port, and to WD Sentinel licensing servers.
Attacks on mail services
In Q4 2015, mail services were especially popular with DDoS attackers.
In particular, activity was detected by the Armada Collective cybercriminal group, which uses DDoS attacks to extort money from its victims. The group is suspected of being involved in an attack on the ProtonMail secure e-mail service in which the cybercriminals demanded $6000 to end the DDoS attack.
In Q4 2015, the largest numbers of DDoS attacks targeted victims in China, the US and South Korea. #KLReportTweet
Statistics for botnet-assisted DDoS attacks
Kaspersky Lab has extensive experience in combating cyber threats, including DDoS attacks of various types and levels of complexity. The company’s experts monitor botnet activity with the help of the DDoS Intelligence system.
The DDoS Intelligence system (part of Kaspersky DDoS Protection) is designed to intercept and analyze commands sent to bots from command and control (C&C) servers, and does not have to wait until user devices are infected or cybercriminal commands are executed in order to gather data.
This report contains the DDoS Intelligence statistics for the fourth quarter of 2015.
In the context of this report, a single (separate) DDoS attack is defined as an incident during which any break in botnet activity lasts less than 24 hours. If the same web resource was attacked by the same botnet after a break of more than 24 hours, this is regarded as a separate DDoS attack. Attacks on the same web resource from two different botnets are also regarded as separate attacks.
The geographic distribution of DDoS victims and C&C servers is determined according to their IP addresses. In this report, the number of DDoS targets is calculated based on the number of unique IP addresses reported in the quarterly statistics.
It is important to note that DDoS Intelligence statistics are limited to those botnets that were detected and analyzed by Kaspersky Lab. It should also be highlighted that botnets are just one of the tools used to carry out DDoS attacks; therefore, the data presented in this report does not cover every DDoS attack that has occurred within the specified time period.
- In Q4, resources in 69 countries were targeted by DDoS attacks.
- 94.9% of the targeted resources were located in 10 countries.
- The largest numbers of DDoS attacks targeted victims in China, the US and South Korea.
- The longest DDoS attack in Q4 2015 lasted for 371 hours (or 15.5 days).
- SYN DDoS, TCP DDoS and HTTP DDoS remain the most common DDoS attack scenarios.
- The popularity of Linux-based bots continued to grow: the proportion of DDoS attacks from Linux-based botnets in the fourth quarter was 54.8%.
Geography of attacks
By the end of 2015, the geography of DDoS attacks narrowed to 69 countries. 94.9% of targeted resources were located in 10 countries.
Q4 saw a considerable increase in the proportion of DDoS attacks targeting resources located in China (from 34.5% to 50.3%) and South Korea (from 17.7% to 23.2%).
Distribution of unique DDoS attack targets by country, Q3 vs Q4 2015
The share of DDoS targets located in the US dropped by 8 percentage points, which saw it move down to third place and South Korea climb to second.
Croatia with 0.3% (-2.5 percentage points) and France, whose share fell from 1.1% to 0.7%, left the Top 10. They were replaced by Hong Kong, with the same proportion as the previous quarter, and Taiwan, whose share increased by 0.5 percentage points.
The statistics show that 94% of all attacks had targets within the Top 10 countries:
Distribution of DDoS attack by country, Q3 vs Q4 2015
In the fourth quarter, the Top 3 ranking remained the same, although the US and South Korea swapped places: South Korea’s contribution grew by 4.3 percentage points, while the US share dropped by 11.5 percentage points. The biggest increase in the proportion of DDoS attacks in Q4 was observed in China – its share grew by 18.2 percentage points.
Changes in DDoS attack numbers
In Q4 2015, DDoS activity was distributed more or less evenly, with the exception of one peak that fell in late October and an increase in activity in early November.
The peak number of attacks in one day was 1,442, recorded on 2 November. The quietest day was 1 October – 163 attacks.
Number of DDoS attacks over time* in Q4 2015.
* DDoS attacks may last for several days. In this timeline, the same attack may be counted several times, i.e. one time for each day of its duration.
Monday and Tuesday were the most active days of the week in terms of DDoS attacks. In Q4, the number of attacks carried out on a Monday was 5.7 percentage points more than in the previous quarter. The figure for Tuesdays changed slightly (-0.3 percentage points).
Distribution of DDoS attack numbers by day of the week, Q4 2015
Types and duration of DDoS attacks
97.5% of DDoS targets in Q4 2015 (vs. 99.3% in Q3) were attacked by bots belonging to one family. In just 2.4% of all cases cybercriminals launched attacks using bots from two different families (used by one or more botnet masters). In 0.1% of cases three or more bots were used, mainly from the Sotdas, Xor and BillGates families.
The longest DDoS attack in Q4 2015 lasted for 371 hours (or 15.5 days). #KLReportTweet
The ranking of the most popular attack methods remained unchanged, although SYN DDoS (57%) and TCP DDoS (21.8%) added 5.4 and 1.9 percentage points respectively.
The distribution of DDoS attacks by type
Once again, most attacks lasted no longer than 24 hours in Q4 2015.
The distribution of DDoS attacks by duration (hours)
The maximum duration of attacks increased again in the fourth quarter. The longest DDoS attack in the previous quarter lasted for 320 hours (13.3 days); in Q4, this record was beaten by an attack that lasted 371 hours (15.5 days).
C&C servers and botnet types
In Q4 2015, South Korea maintained its leadership in terms of the number of C&C servers located on its territory, with its share growing by 2.4 percentage points. The US share decreased slightly – from 12.4% to 11.5%, while China’s contribution grew by 1.4 percentage points.
In Q4 2015, SYN DDoS, TCP DDoS and HTTP DDoS remain the most common DDoS attack scenarios. #KLReportTweet
The Top 3 ranking remained the same. The countries in fourth and fifth switched places – Russia’s share increased from 4.6% to 5.5%, while the share of the UK declined from 4.8% to 2.6%.
Distribution of botnet C&C servers by country in Q4 2015
The proportion of DDoS attacks from Linux-based botnets in Q4 2015 was 54.8% #KLReportTweet
In Q4, the correlation between active bots created from Windows and Linux saw the proportion of attacks by Linux bots grow from 45.6% to 54.8%.
Correlation between attacks launched from Windows and Linux botnets
Events in Q4 2015 demonstrated that the cybercriminals behind DDoS attacks utilize not only what are considered to be classic botnets that include workstations and PCs but also any other vulnerable resources that are available. These include vulnerable web applications, servers and IoT devices. In combination with new channels for carrying out reflection DDoS attacks this suggests that in the near future we can expect a further increase in DDoS capacity and the emergence of botnets consisting of new types of vulnerable devices.