OpenNMS Horizon before 26.0.1 Deserialization of Untrusted Data Vulnerability

Summary

I began to have a look at your software (suite) OpenNMS. Yesterday, I found that one could get Remote Code Execution (RCE) via malicious MQ messages on the Horizon base station from a remote machine with minion credentials. At least that was the case I found quickly to be valid with respect to the role model. Find attached a short write-up describing a little bit more on the exploitation steps.

Credit:

The information has been provided by Jeff Gehlbach

The original article can be found at:https://github.com/OpenNMS/opennms/releases/tag/opennms-26.0.1-1


Details

An issue was discovered in OpenNMS Horizon before 26.0.1, and Meridian before 2018.1.19 and 2019 before 2019.1.7. The ActiveMQ channel configuration allowed for arbitrary deserialization of Java objects (aka ActiveMQ Minion payload deserialization), leading to remote code execution for any authenticated channel user regardless of its assigned permissions.

 

Vulnerable Systems:

OpenNMS Horizon before 26.0.1

OpenNMS Meridian before 2018.1.19

OpenNMS 2019 before 2019.1.7

 

CVE Information:

CVE-2020-12760

 

Disclosure Timeline:
Published Date:5/11/2020

Categories: News