Robust Database Design: CliffGuard
Physical database designs that remain effective when workloads and estimates change.
Research question
Can physical database design remain effective when workload estimates are wrong?
Approach
Model uncertainty explicitly and search for physical designs with robust performance across plausible workloads.
Key results
- CliffGuard treats workload uncertainty as a first-class input to physical database design.
- It searches for index and configuration choices that avoid performance cliffs across plausible future workloads.
- The open-source WorkloadMiner and CliffGuard suite supports workload analysis and robust physical design.
Publications
CliffGuard: A Principled Framework for Finding Robust Database Designs
SIGMOD 2015 · ACM SIGMOD International Conference on Management of Data
Paper ↗Code ↗Project ↗
Cite
@inproceedings{a55995d1-caca-4dbc-a138-c382c0a1e41c,
title = {CliffGuard: A Principled Framework for Finding Robust Database Designs},
author = {Barzan Mozafari and Eugene Zhen Ye Goh and Dong Young Yoon},
booktitle = {ACM SIGMOD International Conference on Management of Data},
year = {2015},
doi = {10.1145/2723372.2749454}
}