Construction of Agent-Based Trust in Cloud Infrastructure
Johannes Harungguan Sianipar, Eyad Saleh, Christoph Meinel · 2014
By design, the cloud system does not allow a cloud administrator to access the customer data in a virtual machine (VM) without customer's knowledge. However, a cloud administrator is able to modify the software/hardware configuration in a way that allow unauthorized access to the customer data. This is because the cloud administrator has full control of the cloud infrastructure. He is a super user in the cloud system and has physical access on the cloud infrastructure. We introduce the ABTiCI (Agent-Based Trust in Cloud Infrastructure) system to detect unauthorized access by verifying and monitoring the Integrity of cloud infrastructure security relevant parts. ABTiCI performs integrity verification at boot-time and at run-time. ABTiCI uses trusted boot with TPM (Trusted Platform Module) to perform integrity verification at boot-time. ABTiCI also monitors access to security relevant parts, such as hardware/software configuration, to be able to detect any changes at run-time. ABTiCI uses agents to do the integrity verification and to communicate between entities in the cloud infrastructure. ABTiCI informs the Certifier about the Dom0 address of the customer VMs (Virtual Machines) to be able to verify whether an integrity verification agent is installed and running in every Dom0.