SUMMARY

As the complexity of modern storage solutions and data-centric applications grows, ensuring their efficiency, resiliency, and security becomes increasingly challenging.

Traditional benchmarking and observability tools often fail to thoroughly characterize and assess the intricate aspects of these advanced systems. This gap makes it difficult to identify weaknesses, optimize performance, and guarantee data integrity and security in these increasingly complex environments.

Our research in the domain of Benchmarking and Observability aims to address the aforementioned challenges by developing robust tools and methodologies specifically tailored for modern storage solutions and data-centric applications.

OBJECTIVES

Within this domain, our current areas of interest include:

  • Benchmarking: We design benchmarking tools that can accurately evaluate storage systems by providing features such as realistic content generation, storage access patterns, data integrity validation, and fault injection. Our goal is to thoroughly assess storage system performance under various conditions, ensuring they meet the demands of data-centric applications.
  • Fault Injection: We develop advanced fault-injection tools to simulate and reproduce various failure scenarios within storage systems. These tools help identify potential vulnerabilities and assess the resiliency of storage solutions. By understanding how systems behave under fault conditions, we can devise strategies to enhance their robustness and reliability.
  • Observability: We develop scalable and non-intrusive observability and diagnosis solutions tailored to complex data-centric applications and systems. These tools offer in-depth insights into system behavior, helping to understand applications' storage access patterns and diagnose issues that might compromise performance, resiliency, or security.

SELECTED PUBLICATIONS

  1. When Amnesia Strikes: Understanding And Reproducing Data Loss Bugs With Fault InjectionRamos, M., Azevedo, J., Kingsbury, K., Pereira, J., Esteves, T., Macedo, R., & Paulo, J.Proceedings of the VLDB Endowment. 2024.

  1. Toward A Practical And Timely Diagnosis of Application’s I/O BehaviorEsteves, T., Macedo, R., Oliveira, R., & Paulo, J.IEEE Access. 2023.

  1. CaT: Content-Aware Tracing and Analysis for Distributed SystemsEsteves, T., Neves, F., Oliveira, R., & Paulo, J.In 22nd International Middleware Conference. ACM, 2021.

Check the full list of publications here.

RESPONSIBLE FOR THE DOMAIN

Tânia Esteves

Assistant Researcher