This repository contains the source code for "From DAGs to Riches: Improving Transaction Throughput in a Deterministic System", our final project for CMSC 624 (Database System Architecture and Implementation).
DAGER is a novel algorithm designed to enhance transaction throughput in deterministic database systems. It utilizes directed acyclic graphs (DAGs) for efficient transaction scheduling to improve transaction throughput. We implement DAGER within an in-memory key-value database (somewhat based on Assignment 2), simulating transactions under varying contention levels.
a2/– Continuous and epoch-based implementations of DAGERpython_graph/– Experimenting with DAG algorithms (not part of the final project)
Pranav Sivaraman • Rakrish Dhakal • Alex Movsesyan • Sathwik Yanamaddi • Daniel Wei