This site is the archived OWASP Foundation Wiki and is no longer accepting Account Requests.
To view the new OWASP Foundation website, please visit https://owasp.org

Testing for Default or Guessable User Account (OWASP-AT-003)

From OWASP
Revision as of 22:42, 10 June 2008 by Khorvath (talk | contribs) (References)

Jump to: navigation, search

[Up]
OWASP Testing Guide v2 Table of Contents

Brief Summary

Today's web application typically runs on popular software, open source or commercial, that is installed on servers and requires configuration or customization by the server administrator. In addition, most of today's hardware appliances, i.e. network routers, database servers, etc., offer web-based configurations or administrative interfaces.

Often, these applications are not properly configured and the default credentials provided for authentication are never updated. In addition, it is typical to find generic accounts such as for testing or administration that use common usernames and passwords are left enabled on the application.
These default username/password combinations are widely known by penetration testers and malicious hackers that can use them to gain access to various types of custom or commercial applications.

Description of the Issue

The sources for this problem are often inexperienced IT personnel, who are unaware of the importance of changing default passwords on installed infrastructure components, programmers, who leave backdoors to easily access and test the application and later forgetting to remove them, application administrators and users that choose an easy username and password for themselves, and applications with built in, non-removable default accounts with a pre-set username and password. Another problem is blank passwords, which are simply a result of security unawareness and a desire to simplify administration.

Black Box testing and example

In blackbox testing we know nothing about the application, its underlying infrastructure, and any username or password policies. Often this is not the case and some information about the application is provided – simply skip the steps that refer to obtaining information you already have.

When testing a known application interface, such as a Cisco router web interface, or Weblogic admin access, check the known usernames and passwords for these devices. This can be done either by Google, or using one of the references in the Further Reading section.

When facing a home-grown application, to which we do not have a list of default and common user accounts, we need to test it manually, following these guidelines:

  • Try the following usernames - "admin", "administrator", "root", "system", "guest", "operator", or "super". These are popular among system administrators and are often used. Additionally you could try "qa", "test", "test1", "testing", "guest" and similar names. Attempt any combination of the above in both the username and the password fields. If the application is vulnerable to username enumeration, and you successfully managed to identify any of the above usernames, attempt passwords in a similar manner. In addition try an empty password or one of the following "password", "pass123", "password123", "admin", or "guest" with the above accounts or any other enumerated accounts.
  • Application administrative users are often named after the application. This means if you are testing an application named "Obscurity", try using obscurity/obscurity or any other similar combination as the username and password.
  • When performing a test for a customer, attempt using names of contacts you have received as usernames with any common passwords.
  • Viewing the User Registration page may help determine the expected format and length of the application usernames and passwords. If a user registration page does not exist, determine if the organization uses a standard naming convention for user names such as their email address or the name before the "@" in the email.
  • Attempt using all the above usernames with blank passwords.
  • Try to extrapolate from the application how usernames are created. such as can a user create his own account or does the system create an account for you based on some personal information or predictable sequence. If the application does create is own account in a predictable sequence, such as user7811, try fuzzing all possible accounts recursively. If you can identify a different response from the application with a valid username and bad password then you can try a brute force attack on the valid username (or quickly try any of the identified common passwords above or in the reference section).
  • If the application creates its own passwords for new users, whether or not the username is created by the application or by the user, then try to determine if the password is predictable. Try to create as many new accounts in quick succession to compare and determine if the passwords are predictable. If predictable then try to correlate these with the usernames, or any enumerated accounts, and use them as a basis for a brute force attack.

Result Expected:
Authorized access to system or application being tested.

Gray Box testing and example

The steps described next rely on an entirely Gray Box approach. If only some of the information is available to you, refer to black box testing to fill the gaps.

Talk to the IT personnel to determine which passwords they use for administrative access.

Check whether these usernames and passwords are complex, difficult to guess, and not related to the application name, person name, or administrative names ("system"). Note blank passwords. Check in the user database for default names, application names, and easily guessed names as described in the Black Box testing section. Check for empty password fields.

Examine the code for hard coded usernames and passwords.

Check for configuration files that contain usernames and passwords.

Result Expected:
Authorized access to system being tested

References

Whitepapers



OWASP Testing Guide v2

Here is the OWASP Testing Guide v2 Table of Contents