TuningChef: an approach for choosing the best cost-benefit database tuning actions

• Victor Augusto Lima Lins de Souza Pontifícia Universidade Católica do Rio de Janeiro (PUC-Rio)
• Sérgio Lifschitz Pontifícia Universidade Católica do Rio de Janeiro (PUC-Rio)

DOI: https://doi.org/10.5753/sbbd.2022.226196

Resumo

While many research works propose a way to list a set of fine-tuning options for a given workload, only a few offer a way to help the DBA make better decisions when encountering a set of available options, especially when taking his possibilities into consideration. We propose and develop a step-by-step decision process. Given a set of fine-tuning options, we recommend the most cost-benefit subset. Enough context for the DBA accompanies the recommendation to understand its reasoning, with the possibility of letting the user build his own subset and check the expected impact. We show some experimental results on actual database systems that further explain our approach and solution.

Palavras-chave: RDBMS, tuning, database, materialized view

Referências

Aken, D. V., Pavlo, A., Gordon, G. J., and Zhang, B. (2017). Automatic database management system tuning through large-scale machine learning. In Proceedings of the 2017 ACM International Conference on Management of Data, pages 1009-1024. ACM.

Ding, B., Das, S., Marcus, R., Wu, W., Chaudhuri, S., and Narasayya, V. R. (2019). AI meets AI: Leveraging query executions to improve index recommendations. In Procs Intl Conf on Management of Data, pages 1241-1258, Amsterdam Netherlands.

Kossmann, J., Halfpap, S., Jankrift, M., and Schlosser, R. (2020). Magic mirror in my hand, which is the best in the land? an experimental evaluation of index selection algorithms. Proc. VLDB Endow., 13(12):2382-2395.

Oliveira, R. (2015). Ontology-based fine tuning: the case of materialized views (in portuguese). Master’s thesis, PUC-Rio.

Oliveira, R. (2019). Automatic Selection and Combination of Tuning Actions (in portuguese). Phd, PUC-Rio.

Perciliano, L., dos V. Santos, Baião, F., Haeusler, E. H., Lifschitz, S., and Almeida, A. C. (2021). Inferencing relational database tuning actions with ondbtuning ontology. In Anais do XXXVI Simp. Bras. de Bancos de Dados (SBBD), pages 157-168.

Schlosser, R. and Halfpap, S. (2020). A decomposition approach for risk-averse index selection. In 32nd Intl Conf Scientific and Statistical Database Management.

Souza, V. (2022). TuningChef: an approach for choosing best cost-benefit tuning actions (in portuguese). Msc, PUC-Rio.

Trummer, I. (2022). DB-BERT: A database tuning tool that reads the manual. In Procs Intl Conf on Management of Data, pages 190-203.

Zhang, J., Zhou, K., Li, G., Liu, Y., Xie, M., Cheng, B., and Xing, J. (2021). CDBTune+: An efficient deep reinforcement learning-based automatic cloud database tuning system. The VLDB Journal, 30(6):959-987.