Implementing a File Architecture for a Database Operating System

Widely used operating systems such as Linux are becoming outdated. Because they were optimized for the limited processing power of several decades ago, scalability is a growing concern given the powerful computing environments available now. Instead of adding onto the current operating system design to address these problems, our team proposed a design for a system we call the Database Operating System (DBOS) that uses database tables to represent the state, and queries to represent operations to the state. In this study, I show that the performance of this new OS design is competitive with current operating systems. In order to obtain performance metrics, the following few key components are in focus: the file system, scheduler, and IPC handler. This study focuses on the file system implementation. The file system is a simple file architecture using VoltDB as our in-memory database, with tables representing files and stored procedures representing IO tasks such as read and write. DBOS uses main memory as the primary storage, and mechanisms were implemented to spill data to disk when necessary. Benchmark tests were conducted against DBOS and other existing operating systems, which prove that DBOS is not just competitive with, but can outperform, existing operating systems.

Name thesis:degree_name

Master

Department dc:contributor.department

Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science

Grantor dc:publisher

Massachusetts Institute of Technology

Year dc:date.issued

2021

Author dc:creator

• Hong, Daniel

Advisor dc:contributor.advisor

• Stonebraker, Michael

Statement dc:rights

• In Copyright - Educational Use Permitted
• Copyright MIT

Licence dc:rights.uri

• http://rightsstatements.org/page/InC-EDU/1.0/

Handle dc:identifier.uri

https://hdl.handle.net/1721.1/139200

OAI identifier oai:identifier

oai:dspace.mit.edu:1721.1/139200