ProtoGENI security: threats to resources and run-time interactions

Abstract

The ever extending threats to Internet community, and financial, physical, mental, social damages because of it forced researchers to rethink about the architecture, components, and services of future Internet. Security was not the concern at the time of development of existing architecture of Internet There is a very high probability to attack current Internet without being caught which supports the proliferation of cyber crimes. Security is one of the prime objectives of future Internet which is highly obscure term. This is the challenge for ever to maintain the security of the Internet as attackers probably have higher intelligence and determination to break the security. GENI is a virtual lab to provide all necessary resources and environment closer to expected future Internet so that researchers can test the innovative ideas to develop a more secure, accountable, usable, and manageable future Internet. ProtoGENI is a prototype of GENI and it is in function to test network research ideas. ProtoGENI requires a rigorous observation and implementation improvements to achieve intended security in GENI. Security of ProtoGENI is crucial as experiment results can provide a false picture of security capabilities if they are being tested in an environment which can be manipulated by malicious users or not consistent in its performance. It can affect the security of whole system drastically, and destroying the whole effort of developing a secure future Internet. This work is an effort to test and observe the existing security mechanism and functioning of the ProtoGENI system, and to find out the exploitable attacking loophole. The initial experiments, results, and observations provide a detailed functioning and security problems which can be utilized to improve the overall ProtoGENI security architecture. Though Security is a process and not a product, this work is to provide the current security issues and suggestions to improve security settings involving all components which work together to utilize ProtoGENI facilities for testing innovative ideas for developing future Internet. Threats to ProtoGENI resources and runtime interactions are in focus for this research work. It explores the existing functioning and possible security weaknesses to cause a non-functional, semi non-functional or malfunctioned system. There are many observations during executing experiments which affect the performance of the system. These observations can assist to improve the overall ProtoGENI functionality. Results indicate that there are threats to resources and run-time interactions between ProtoGENI components. Non-availability and non-usability of resources can affect the network experiments severely. Cross-experiment communication is also possible in wireless Emulab experiments. Initial Wireless communication analysis on Emulab provides details of wireless traffic behavior and traffic interferences. Overall security at host machine can be enhanced by modifying default security settings including SSH port number and root login rights. An alternative solution is provided to solve default XMLPC server settings to establish initial setup for executing ProtoGENI experiments. These findings are subject to time-line to the progress of ProtoGENI and GENI projects. This work can assist novice ProtoGENI researchers to understand the basic functionality, associated problems, and possible solutions. These initial findings for security issues in existing ProtoGENI system and observations will assist to improve the overall security functionality of ProtoGENI.