Unpatched Apache reverse proxy flaw allows access to internal network
- 26 November, 2011 00:43
- Comments
A yet-to-be-patched flaw discovered in the Apache HTTP server allows attackers to access protected resources on the internal network if some rewrite rules are not defined properly.
The vulnerability affects Apache installations that operate in reverse proxy mode, a type of configuration used for load balancing, caching and other operations that involve the distribution of resources over multiple servers.
In order to set up Apache HTTPD to run as a reverse proxy, server administrators use specialized modules like mod_proxy and mod_rewrite.
Security researchers from Qualys warn that if certain rules are not configured correctly, attackers can trick servers into performing unauthorized requests to access internal resources.
The problem isn't new and a vulnerability that allowed similar attacks was addressed back in October. However, while reviewing the patch for it, Qualys researcher Prutha Parikh realized that it can be bypassed due to a bug in the procedure for URI (Uniform Resource Identifier) scheme stripping. The scheme is the URI part that comes before the colon ":" character, such as http, ftp or file.
One relatively common rewrite and proxying rule is "^(.*) http://internal_host$1", which redirects the request to the machine internal_host. However, if this is used and the server receives, for example, a request for "host::port" (with two colons), the "host:" part is stripped and the rest is appended to http://internal_host in order to forward it internally.
The problem is that in this case, the remaining part is ":port", therefore transforming the forwarded request into http://internal_host:port, an unintended behavior that can result in the exposure of a protected resource.
In order to mitigate the problem server administrators should add a forward slash before $1 in the rewrite rule, the correct form being "^(.*) http://internal_host/$1", Parikh said.
The Apache developers are aware of the problem and are currently discussing the best method of fixing it. One possibility would be to strengthen the previous patch in the server code in order to reject such requests, however, there's no certainty that other bypass methods won't be discovered.
"We could try improve that fix, but I think it would be simpler to change the translate_name hooks in mod_proxy and mod_rewrite to enforce the requirement in the 'right' place," said Red Hat senior software engineer Joe Orton on the Apache dev mailing list. Orton proposed a patch that is currently being reviewed by the other developers.
Join the CIO Australia group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.
- Bookmark this page
- Share this article
- Got more on this story? Email CIO
- Follow CIO on twitter
- In Pictures: Google I/O 2013's coolest products and services
- Report: Yahoo board approves deal to buy Tumblr for $US1.1bn
- Dell's thumb PC, Project Ophelia, to ship in July
- IT doesn’t see much value in making Windows 8 enterprise standard: Forrester
- BT Financial embraces BI to better understand customers
-
Spiceworks' free management software gets integrated MDM
-
Opinion: Why national e-health is not for everyone
-
Opinion: Why national e-health is not for everyone
-
Opinion: Why national e-health is not for everyone
-
Opinion: Why national e-health is not for everyone
-
Managing the Rapid Rise in Database Growth: 2011 IOUG Survey on Database Manageability
As the era of “Big Data” marches on unabated, data is coming from an ever wider range of sources, including transactional systems, mobile devices, sensors, streaming media, and social networks. Businesses are looking for innovative ways to better leverage terabytes—and for some, petabytes—of information. Read more. -
HP Helps NEC Reduce Network Management Costs and Gain Efficiencies
NEC wanted to reduce network management costs, while increasing network visibility, decreasing mean-time-to-repair, improving stability and mitigating the risk of downtime. Download today to hear from Cameron Craig, Senior department manager of NEC on what approach they took and why. -
IDC: Delivering Customer Value with Enterprise Flash Deployments
When it comes to flash, “one size does not fit all.” IDC examines recent flash trends in enterprise storage deployments. This includes: highlighting how SSDs are filling in gaps of existing storage systems when coupled with intelligent archiving and automated tiering, the pros and cons of different SSD approaches, and tips to overcome concerns of reliability, manageability and scalability.
















