Abstract
Indexes are vital for data retrieval performance. For online scenarios with dynamic workloads, index tuning is challenging. A commonly used strategy is to launch tuning requests periodically, yet resource-intensive tuning sessions can obstruct it, particularly when dealing with frequently varying workloads.
To tackle this challenge, we propose a learned tuning diagnostic that can be incorporated into the Monitor-Diagnose-Tune paradigm for online index tuning. Rather than invoking a comprehensive tuning tool every time a triggering condition occurs, the tuning diagnostic serves to determine whether a tuning session should be launched. By formulating the determination of sub-optimal index configurations as a classification task in machine learning, our approach can effectively identify whether the current index configuration is sub-optimal. To circumvent the need for costly data collection for each database instance, we propose a transferable representation of queries and indexes that allows for cross-database learning. Our comprehensive empirical results on the TPC-H and TPC-DS benchmarks demonstrate that our approach can reduce the total time by up to 13.3% and the number of optimizer what-if calls by up to 36% compared to the baselines, and validate the effectiveness of our transferable representation in cross-database learning.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
HypoPG - Hypothetical Indexes for PostgreSQL. https://github.com/HypoPG/hypopg
TPC-DS benchmark. www.tpc.org/tpcds
TPC-H benchmark. www.tpc.org/tpch
Agrawal, S., Chu, E., Narasayya, V.R.: Automatic physical design tuning: workload as a sequence. In: Proceedings of the 2006 ACM SIGMOD International Conference on Management of Data (2006)
Bruno, N., Chaudhuri, S.: Automatic physical database tuning: a relaxation-based approach. In: ACM SIGMOD Conference (2005)
Bruno, N., Chaudhuri, S.: To tune or not to tune?: a lightweight physical design alerter. In: Very Large Data Bases Conference (2006)
Bruno, N., Chaudhuri, S.: An online approach to physical design tuning. In: 2007 IEEE 23rd International Conference on Data Engineering, pp. 826–835 (2007)
Chaudhuri, S., Narasayya, V.: AutoAdmin “what-if” index analysis utility. ACM SIGMOD Rec. 27(2), 367–378 (1998)
Chaudhuri, S., Narasayya, V.R.: An efficient cost-driven index selection tool for Microsoft SQL server. In: Very Large Data Bases Conference (1997)
Choenni, S., Blanken, H.M., Chang, T.: Index selection in relational databases. In: Proceedings of ICCI 1993: 5th International Conference on Computing and Information, pp. 491–496 (1993)
Dash, D., Polyzotis, N., Ailamaki, A.: Cophy: a scalable, portable, and interactive index advisor for large workloads. arXiv abs/1104.3214 (2011)
Deep, S., Gruenheid, A., Koutris, P., Naughton, J.F., Viglas, S.: Comprehensive and efficient workload compression. arXiv abs/2011.05549 (2020)
Ding, B., Das, S., Marcus, R., Wu, W., Chaudhuri, S., Narasayya, V.R.: AI meets AI: leveraging query executions to improve index recommendations. In: Proceedings of the 2019 International Conference on Management of Data (2019)
Gao, J., Zhao, N., Wang, N., Hao, S., Wu, H.: Automatic index selection with learned cost estimator. Inf. Sci. 612, 706–723 (2022)
Hilprecht, B., Binnig, C.: Zero-shot cost models for out-of-the-box learned cost prediction. Proc. VLDB Endow. 15, 2361–2374 (2022)
Jain, S., Howe, B., Yan, J., Cruanes, T.: Query2vec: an evaluation of NLP techniques for generalized workload analytics. arXiv, Databases (2018)
Kipf, A., Kipf, T., Radke, B., Leis, V., Boncz, P.A., Kemper, A.: Learned cardinalities: estimating correlated joins with deep learning. arXiv abs/1809.00677 (2018)
Kossmann, J., Halfpap, S., Jankrift, M., Schlosser, R.: Magic mirror in my hand, which is the best in the land? Proc. VLDB Endow. 13, 2382–2395 (2020)
Kossmann, J., Kastius, A., Schlosser, R.: SWIRL: selection of workload-aware indexes using reinforcement learning. In: International Conference on Extending Database Technology (2022)
Lan, H., Bao, Z., Peng, Y.: An index advisor using deep reinforcement learning. In: Proceedings of the 29th ACM International Conference on Information & Knowledge Management (2020)
Li, G., Zhou, X., Li, S., Gao, B.: Qtune: a query-aware database tuning system with deep reinforcement learning. Proc. VLDB Endow. 12, 2118–2130 (2019)
Licks, G.P., Couto, J.C., de Fátima Miehe, P., de Paris, R., Ruiz, D.D., Meneguzzi, F.: SmartIX: a database indexing agent based on reinforcement learning. Appl. Intell. 50, 2575–2588 (2020)
Ma, L., Aken, D.V., Hefny, A.S., Mezerhane, G., Pavlo, A., Gordon, G.J.: Query-based workload forecasting for self-driving database management systems. In: Proceedings of the 2018 International Conference on Management of Data (2018)
Paul, D., Cao, J., Li, F., Srikumar, V.: Database workload characterization with query plan encoders. Proc. VLDB Endow. 15, 923–935 (2021)
Sadri, Z., Gruenwald, L., Leal, E.: DRLindex: deep reinforcement learning index advisor for a cluster database. In: Proceedings of the 24th Symposium on International Database Engineering & Applications (2020)
Schlichtkrull, M., Kipf, T., Bloem, P., van den Berg, R., Titov, I., Welling, M.: Modeling relational data with graph convolutional networks. In: Extended Semantic Web Conference (2017)
Schlosser, R., Kossmann, J., Boissier, M.: Efficient scalable multi-attribute index selection using recursive strategies. In: 2019 IEEE 35th International Conference on Data Engineering (ICDE), pp. 1238–1249 (2019)
Schnaitter, K., Abiteboul, S., Milo, T., Polyzotis, N.: On-line index selection for shifting workloads. In: 2007 IEEE 23rd International Conference on Data Engineering Workshop, pp. 459–468 (2007)
Sharma, A.K., Schuhknecht, F.M., Dittrich, J.: The case for automatic database administration using deep reinforcement learning. arXiv abs/1801.05643 (2018)
Sun, J., Li, G.: An end-to-end learning-based cost estimator. arXiv abs/1906.02560 (2019)
Surajit, C., Narasayya, V.R.: Anytime algorithm of database tuning advisor for Microsoft SQL server (2020)
Tang, X., Wu, S., Song, M., Ying, S., Li, F.Y., Chen, G.: PreQR: pre-training representation for SQL understanding. In: Proceedings of the 2022 International Conference on Management of Data (2022)
Valavala, M., Alhamdani, W.: Automatic database index tuning using machine learning. In: 2021 6th International Conference on Inventive Computation Technologies (ICICT), pp. 523–530. IEEE (2021)
Wu, W., et al.: Budget-aware index tuning with reinforcement learning. In: Proceedings of the 2022 International Conference on Management of Data (2022)
Yan, Y., Yao, S., Wang, H., Gao, M.: Index selection for NoSQL database with deep reinforcement learning. Inf. Sci. 561, 20–30 (2021)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Hang, H., Sun, J. (2023). Enhancing Online Index Tuning with a Learned Tuning Diagnostic. In: Strauss, C., Amagasa, T., Kotsis, G., Tjoa, A.M., Khalil, I. (eds) Database and Expert Systems Applications. DEXA 2023. Lecture Notes in Computer Science, vol 14146. Springer, Cham. https://doi.org/10.1007/978-3-031-39847-6_14
Download citation
DOI: https://doi.org/10.1007/978-3-031-39847-6_14
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-39846-9
Online ISBN: 978-3-031-39847-6
eBook Packages: Computer ScienceComputer Science (R0)