About me

I'm a Ph.D. student in the school of Electrical and Computer Engineering (ECE) at Purdue University. I'm associated to the Dependable Computing Systems Lab (DCSL) working under the supervision of Professor Saurabh Bagchi. I completed my M.S. in ECE at Purdue in August, 2008.

I was born in Panama and had my B.S. in Electronics and Communications Engineering from the University of Panama in 2002. I came to Purdue with a fellowship granted by my home-country government in 2006 (SENACYT-IFARHU fellowship).

Before graduate school, I worked for the National Bureau of Science and Technology (SENACYT) of my country during 2005-2006. Before that, I served as a Security Officer at the Latin American Export Bank in 2004.

Research

My research interests lay in the area of error detection and diagnosis in distributed systems. I am particularly interested in answering the question of how to design and implement dependable distributed systems, and how to achieve a good balance between performance, reliability and security. The following is a description of the projects I'm involved on:

Error Detection in Distributed Applications:

From fall 2006, I have been involved in the project "Detection & Diagnosis in Distributed Applications through Non-Intrusive Monitoring". This project has allowed me to study the feasibility of building a system, called Monitor, for detection and diagnosis of failures in distributed computing environments, in an on-line and non-intrusive way. Monitor is able to detect errors at runtime through the use of normal-behaviour rules, given a Finite State Machine (FSM) of the monitored application (or protocol).

Detection in High Data Rate Applications:

Today's distributed applications are composed of multiple components that exchange data (i.e., messages) at a high rate. A detection system gets overwhelmed when it has to process high volume of messages. Since the overall workload of a detection system is proportional to [the number of messages per unit of time] x [the number of rules to match per message], a way to reduce the workload is by reducing the effective messages that the system has to process by sampling them. We have developed sampling techniques whereby some messages are sampled (and the others are dropped) so that a detection system does not reach instability.

Root Cause Analysis in Parallel Systems:

I spent five months at the Livermore Lawrence National Laboratory (LLNL) in 2009 working on a project for providing root cause analysis on MPI applications. I developed a tool that isolates abnormal MPI processes by, first, creating a model of the behaviour of each task and, second, by clustering the models. The tool provides to the user the region of the program where the error is first manifested. This automates the debugging process of large-scale parallel applications. At LLNL I worked in collaboration of Greg Bronevetsky, Bronis de Supinski, Martin Schulz and Dong Ahn.

Publications

• How To Keep Your Head Above Water While Detecting  Errors, Ignacio Laguna, Fahad A. Arshad, David M. Grothe, Saurabh Bagchi, ACM/IFIP/USENIX 10th International Middleware Conference (Middleware 2009), UIUC Illinois, Nov-Dec 2009.

• Scalable Temporal Order Analysis for Large Scale Debugging, Dong H. Ahn, Bronis R. de Supinski, Ignacio. Laguna, Greg L. Lee, Ben Liblit, Barton P. Miller, and Martin Schulz, ACM/IEEE Conference on Supercomputing 2009 (SC 2009), Portland, OR, Nov 2009.

• Distributed Diagnosis of Failures in a Three Tier E-Commerce System, Gunjan Khanna, Ignacio Laguna, Fahad A. Arshad and Saurabh Bagchi, 26th IEEE Symposium on Reliable Distributed Systems (SRDS 2007), Beijing, China, Oct 2007.

• Stateful Detection in High Throughput Distributed Systems, Gunjan Khanna, Ignacio Laguna, Fahad A. Arshad and Saurabh Bagchi, 26th IEEE Symposium on Reliable Distributed Systems (SRDS 2007), Beijing, China, Oct 2007.
  

Professional Experience

• Intern, Lawrence Livermore National Lab, CA, 2009:  Performed research on root cause analysis in parallel programs. Projects: (1) developed a fault injector for emulating dynamically common faults in MPI applications, (2) developed a tool for automatic debugging of parallel tasks.
  

• Research Assistant, ECE Department, Purdue University, IN, 2008-Present: Performed research in the area of dependable computing systems.

• Network Engineer, National Bureau of Science and Technology (SENACYT), Panama, 2005-2006: Managed grant proposals for founding of IT research projects, particulraly in the area of research and education networks.
  

Organizations

• IEEE Member
• USENIX Association Member