About
Researcher & Sfotware Engineer
- Degree: PhD (Expected: 2022)
- Research Focus: Understanding, improving and troubleshooting large-scale distributed systems
As part of my PhD, I have been involved in efforts to:
- Troubleshoot failures in production systems
- Understand and improve fault-tolerance properites in large systems
Work & Education
Sumary
Kamala Ram
email: kamala.ramas@gmail.com
Broadly, I analyze and troubleshoot large distributed systems by posing and answering questions based on observed system executions. For example, to determine the next fault to inject when testing a system for fault tolerance, system designers may ask - What do all successful executions have in common? To troubleshoot an issue, operators may ask - How do unsuccessful executions differ from successful executions? To obtain an overall understanding of the system for feature development, programmers may ask - What can successful executions teach us about how they succeed? In my dissertation research, I answer the three questions posed above by deriving insights from observed system executions (distributed traces and provenance) and building software tools to demonstrate their applicability.
Education
PhD (Computer Science)
2015 - 2022 (Expected)
University of California, Santa Cruz
MSc[Engg] (Computer Science)
2010 - 2013
Indian Institute of Science, Bangalore
B.E. (Computer Science)
2004 - 2008
Visvesvaraya Technological University, Belgaum
Programming Languages
| Skill Level | Programming Languages |
|---|---|
| Proficient | C, Python |
| Familiar | C, C++, Java, Go, TypeScript, Haskell, Perl |
Work Experience
Intern, eBay
2018 - 2022
- Aggregate Comparison of Traces for incident localization
- Understanding fault-tolerance properities of microservices using traces and fault-injection
Intern, Intel Labs
June 2017 - August 2017
- Worked on experimental framework to induce Service Level Aggreement (SLA) violations in GET path for Openstack Swift
Intern, Elastic
July 2016 - September 2016
- Modeled data replication protocol at Elastic (flavor of primary backup)
- For a pre-defined class of faults, our appoach demostrates how we use data lineage to ensure that the expected invariants are upheld even as the system evolves
Software Engineer, Arista Networks
2013 - 2015
- Worked on supporting and enhacing network protocol stack (Programming Language: C)
- Upstreamed a patch to linux kernel to fix support for blackhole and prohibit routes Link to patch
Software Engineer, Hewlett Packard
2008 - 2010
- As an engineer in the Photo team (responsible for displaying thumbnails on-screen to printing of photos), I was involved in feature development and maintenance for a variety of product lines. (Programming language: C)
Publications & Talks
Posters & Publications
Talks & Write-ups
- HPTS, 2019: Automated Fault Diagnostics
- HPTS, 2017: Growing a protocol
- Chaos Community Day, 2017: SLA violations: How and Why?
- Chaos Community Day, 2017: Growing a Protocol
- Blog Post: Model and test data replication at Elasticsearch Link to post