Slowloris (computer security)
Slowloris is a type of denial of service attack tool which allows a single machine to take down another machine's web server with minimal bandwidth and side effects on unrelated services and ports.
Initial release | 17 June 2009 |
---|---|
Stable release | 0.7
|
Written in | Perl |
Platform | Cross-platform |
Size | 36 kb |
Type | Hacking tool |
Website | ha.ckers.org/slowloris/ |
Slowloris tries to keep many connections to the target web server open and hold them open as long as possible. It accomplishes this by opening connections to the target web server and sending a partial request. Periodically, it will send subsequent HTTP headers, adding to, but never completing, the request. Affected servers will keep these connections open, filling their maximum concurrent connection pool, eventually denying additional connection attempts from clients.[1]
The program was named after slow lorises, a group of primates which are known for their slow movement.
Affected web servers
This includes but is not necessarily limited to the following, per the attack's author:[1]
- Apache 1.x and 2.x
- dhttpd
- Websense "block pages" (unconfirmed)
- Trapeze Wireless Web Portal (unconfirmed)
- Verizon's MI424-WR FIOS Cable modem (unconfirmed)
- Verizon's Motorola Set-top box (port 8082 and requires auth - unconfirmed)
- BeeWare WAF (unconfirmed)
- Deny All WAF (patched)[2]
- Flask (development server)
- Internet Information Services (IIS) 6.0 and earlier [3]
- Nginx 1.5.9 and earlier [4]
Vulnerable to Slowloris attack on the TLS handshake process:
Because Slowloris exploits problems handling thousands of connections, the attack has less of an effect on servers that handle large numbers of connections well. Proxying servers and caching accelerators such as Varnish, nginx, and Squid have been recommended[7] to mitigate this particular kind of attack. In addition, certain servers are more resilient to the attack by way of their design, including Hiawatha,[8] IIS, lighttpd, Cherokee, and Cisco CSS.
Mitigating the Slowloris attack
While there are no reliable configurations of the affected web servers that will prevent the Slowloris attack, there are ways to mitigate or reduce the impact of such an attack. In general, these involve increasing the maximum number of clients the server will allow, limiting the number of connections a single IP address is allowed to make, imposing restrictions on the minimum transfer speed a connection is allowed to have, and restricting the length of time a client is allowed to stay connected.
In the Apache web server, a number of modules can be used to limit the damage caused by the Slowloris attack; the Apache modules mod_limitipconn, mod_qos, mod_evasive, mod security, mod_noloris, and mod_antiloris have all been suggested as means of reducing the likelihood of a successful Slowloris attack.[1][9] Since Apache 2.2.15, Apache ships the module mod_reqtimeout as the official solution supported by the developers.[10]
Other mitigating techniques involve setting up reverse proxies, firewalls, load balancers or content switches.[11] Administrators could also change the affected web server to software that is unaffected by this form of attack. For example, lighttpd and nginx do not succumb to this specific attack.[1]
Notable usage
During the protests that erupted in the wake of the 2009 Iranian presidential election, Slowloris arose as a prominent tool used to leverage DoS attacks against sites run by the Iranian government.[12] The belief was that flooding DDoS attacks would affect internet access for the government and protesters equally, due to the significant bandwidth they can consume. The Slowloris attack was chosen instead, because of its high impact and relatively low bandwidth.[13] A number of government-run sites were targeted during these attacks, including gerdab.ir, leader.ir, and president.ir.[14]
A variant of this attack was used by spam network River City Media to force Gmail servers to send thousands of messages in bulk, by opening thousands of connections to the Gmail API with message sending requests, then completing them all at once.[15]
Similar software
Since its release, a number of programs have appeared that mimic the function of Slowloris while providing additional functionality, or running in different environments:[16]
- PyLoris – A protocol-agnostic Python implementation supporting Tor and SOCKS proxies.[17]
- Slowloris – A Python 3 implementation of Slowloris with SOCKS proxy support.[18]
- Goloris – Slowloris for nginx, written in Go.[19]
- slowloris - Distributed Golang implementation[20]
- QSlowloris – An executable form of Slowloris designed to run on Windows, featuring a Qt front end.[21]
- An unnamed PHP version which can be run from a HTTP server.[22]
- SlowHTTPTest – A highly configurable slow attacks simulator, written in C++.[23][24]
- SlowlorisChecker – A Slowloris and Slow POST POC (Proof of concept). Written in Ruby.[25]
- Cyphon - Slowloris for Mac OS X, written in Objective-C.[26]
- sloww - Slowloris implementation written in Node.js.[27]
- dotloris - Slowloris written in .NET Core[28]
- SlowDroid - An enhanced version of Slowloris written in Java, reducing at minimum the attack bandwidth[29]
See also
References
- "Slowloris HTTP DoS". Archived from the original on 26 April 2015. Retrieved 26 June 2009.
{{cite web}}
: CS1 maint: bot: original URL status unknown (link) - "Archived copy" (PDF). Archived from the original (PDF) on 1 February 2014. Retrieved 15 May 2013.
{{cite web}}
: CS1 maint: archived copy as title (link) - "Slowloris". www.powerwaf.com. Retrieved 17 July 2023.
- "Slowloris". www.powerwaf.com. Retrieved 17 July 2023.
- "Slowloris". www.powerwaf.com. Retrieved 17 July 2023.
- "Slowloris". www.powerwaf.com. Retrieved 17 July 2023.
- "How to best defend against a "slowloris" DOS attack against an Apache web server?". serverfault.com. Retrieved 28 December 2016.
- "Performance testing while under attack". hiawatha-webserver.org. 28 February 2014.
- "mod_noloris: defending against DoS". niq's soapbox. July 2009. Retrieved 7 January 2012.
- "mod_reqtimeout - Apache HTTP Server". Httpd.apache.org. Retrieved 3 July 2013.
- Breedijk, Frank (22 June 2009). "Slowloris and Nkiller2 vs. the Cisco CSS load balancer". Cupfighter.net. Archived from the original on 15 February 2012. Retrieved 7 January 2012.
- Zdrnja, Bojan (23 June 2009). "ISC Diary | Slowloris and Iranian DDoS attacks". Isc.sans.org. Retrieved 7 January 2012.
- Archived 29 June 2009 at the Wayback Machine
- Archived 11 August 2009 at the Wayback Machine
- Vickery, Chris (6 March 2017). "Spammergate: The Fall of an Empire". MacKeeper Security Watch. Archived from the original on 6 March 2017.
- Robert "RSnake" Hansen. "Slowloris" (PDF). SecTheory. Retrieved 7 January 2012.
- "PyLoris". MotomaSTYLE. 19 June 2009. Archived from the original on 15 July 2009. Retrieved 7 January 2012.
- "Slowloris rewrite in Python". GitHub. Retrieved 10 May 2017.
- valyala. "Slowloris for nginx DoS". GitHub. Retrieved 4 February 2014.
- Tsankov, Ivaylo (22 April 2022), slowloris - Golang distributed Slowloris attack, retrieved 24 April 2022
- "How to help take down gerdab.ir in 5 easy steps". cyberwar4iran. 28 June 2009. Retrieved 7 January 2012.
- "Full Disclosure: apache and squid dos". Seclists.org. 19 June 2009. Retrieved 7 January 2012.
- "Testing Web Servers for Slow HTTP Attacks". qualys.com. 19 September 2011. Retrieved 13 January 2012.
- "shekyan/slowhttptest: Application Layer DoS attack simulator". GitHub. Retrieved 19 April 2017.
- "Simple script to check if some server could be affected by Slowloris attack". github.com/felmoltor. 31 December 2012. Retrieved 31 December 2012.
- abilash. "Slowloris for OSX". GitHub. Retrieved 8 April 2017.
- Davis, Ethan (17 February 2018), sloww: Lightweight Slowloris attack CLI in Node, retrieved 18 February 2018
- Bassel Shmali (28 November 2021). "Slowloris written in .Net core". GitHub.
- Cambiaso, Enrico; Papaleo, Gianluca; Aiello, Maurizio (2014). "SlowDroid: Turning a Smartphone into a Mobile Attack Vector". 2014 International Conference on Future Internet of Things and Cloud. pp. 405–410. doi:10.1109/FiCloud.2014.72. ISBN 978-1-4799-4357-9. S2CID 14792419.