Publication Date
Fall 2025
Degree Type
Master's Project
Degree Name
Master of Science in Computer Science (MSCS)
Department
Computer Science
First Advisor
Suneuy Kim
Second Advisor
Robert Chun
Third Advisor
Thomas Austin
Keywords
HTAP Benchmark, OLTP, OLAP, GRAPH centric, healthcare, graph databases
Abstract
This thesis presents GraphHTAP-Bench, an HTAP benchmark for graph databases on Electronic Health Record workloads. Using the ORBDA dataset, we build a property-graph schema of longitudinal patient journeys and derive a mixed workload that combines OLTP transactions, relational-style OLAP, graph-centric analytics, and heartbeat-based freshness probes. We run Neo4j, Memgraph, and Janus- Graph under controlled arrival patterns and measure throughput, latency distributions (including tails), and commit-to-visibility freshness. The results show how scan- and traversal-heavy analytics interfere with transactional performance, how clustering changes latency and headroom, and how GraphHTAP-Bench provides reproducible guidance for selecting and tuning graph engines for EHR-style HTAP deployments.
Recommended Citation
Le, Phuoc Dai, "GraphHTAP-Bench - HTAP Benchmarking for Graph Databases" (2025). Master's Projects. 1605.
DOI: https://doi.org/10.31979/etd.grx5-fngc
https://scholarworks.sjsu.edu/etd_projects/1605
