M. Alex Merenstein

Ph.D. Candidate
File Systems and Storage Lab
Advisor: Professor Erez Zadok
Email: mmerenstein@cs.stonybrook.edu
CV

I'm a third year Computer Science Ph.D. student at Stony Brook University. My research interests include operating systems, file systems, networking, and security.

Research Projects

My primary research projects are related to cloud native storage and are collaborations with researchers at IBM Research at Almaden.

File Systems for FaaS
Function as a Service (FaaS, sometimes also known as "serverless computing") platforms have become popular for web applications due to their elasticity and attractive billing and programming models. These features make FaaS potentially useful for non-web applications as well. However, the restrictions currently imposed by the major cloud providers limits the range of applications that can be transitioned to FaaS platforms. The restriction that we are addressing in this research is the lack of access to shared, persistent file systems from FaaS applications. We plan to demonstrate how access to shared persistent file systems enables a wider range of applications to take advantage of FaaS platforms. We also intend to develop a file system that accelerates workloads for common FaaS data exchange scenarios.

Cloud Native Storage Benchmarking
Modern cloud computing is undergoing a transformation, driven by containerization technologies such as Docker and Kubernetes and the rising popularity of microservice based architectures. The result is that cloud workloads are now more diverse and cloud applications are more elastic and more dynamic. Storage operations that were previously infrequent like volume creation and deletion are now executed dozens or hundreds of times every day.

Understanding the performance of an application or storage system in this environment is difficult because there are no tools that can recreate all of these new cloud native conditions. To address this problem, we are designing and building a benchmark capable of generating workloads that are representative of modern, cloud native environments. This benchmark will include measurement and analysis components to help debug performance problems, and it will be used to evaluate and study storage performance in cloud native environments.

Publications

CNSBench: A Cloud Native Storage Benchmark
Alex Merenstein, Vasily Tarasov, Ali Anwar, Deepavali Bhagwat, Julie Lee, Lukas Rupprecht, Dimitris Skourtis, Yang Yang, Erez Zadok
19th USENIX Conference on File and Storage Technologies (FAST '21)
PDF Presentation

The Case for Benchmarking Control Operations in Cloud Native Storage
Alex Merenstein, Vasily Tarasov, Ali Anwar, Deepavali Bhagwat, Lukas Rupprecht, Dimitris Skourtis, Erez Zadok
12th USENIX Workshop on Hot Topics in Storage (HotStorage '20)
PDF Presentation

SpecNFS: A Challenge Dataset Towards Extracting Formal Models from Natural Language Specifications
Sayontan Ghosh, Amanpreet Singh, Alex Merenstein, Wei Su, Scott Smolka, Erez Zadok, Niranjan Balasubramanian
To be presented at The First Workshop on Natural Language Processing for Programming (NLP4Prog 2021)


Education

Work Experience

Teaching Experience