Attacking the Messenger: Exploring the Security of Big Data Messenger Apache Kafka
View/ Open
Abstract
As technology becomes faster, cheaper, and more compact, a higher volume of data must be processed. This demand drives the need to process high volumes of data in near real time. As a result, technologies such as Kafka have been created as high throughput messaging bus systems. Utilizing these new technologies could vastly improve the way we look at data processing, especially when that data is coming from IoT or distributed systems.
Kafka provides multiple methods of encryption and authentication with its brokers, however, securing the producers and consumers is the responsibility of the application owner. This paper focuses on this key aspect in order to consider how an attacker could exploit and compromise the flow and integrity of the data. After access to the producers and consumers has been compromised, examples of data manipulation are performed in order to demonstrate real world consequence of breaches such as these.