pith. machine review for the scientific record. sign in

arxiv: 2604.14725 · v1 · submitted 2026-04-16 · 💻 cs.DB · cs.LG

Recognition: unknown

RELOAD: A Robust and Efficient Learned Query Optimizer for Database Systems

Jaeyoung Sim, Kwanghyun Park, Seokwon Lee, Sihyun Kim, Yiwen Zhu, Yuhsing Li

Authors on Pith no claims yet

Pith reviewed 2026-05-10 10:05 UTC · model grok-4.3

classification 💻 cs.DB cs.LG
keywords query optimizationreinforcement learningdatabase systemslearned optimizersrobustnessefficiencyquery planningperformance regression
0
0 comments X

The pith

RELOAD is a learned query optimizer that reduces individual query performance regressions and reaches expert plan quality faster than prior reinforcement learning methods.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper presents RELOAD to overcome two practical problems with reinforcement learning for database query optimization: unstable results where some queries run much slower than expected, and slow training that fails to match traditional expert optimizers quickly. It targets robustness by cutting query-level regressions and keeping behavior consistent across runs, while improving efficiency through quicker convergence to high-quality plans. Experiments across Join Order Benchmark, TPC-DS, and Star Schema Benchmark show gains of up to 2.4 times in robustness and 3.1 times in efficiency over existing RL techniques. A sympathetic reader would care because these fixes directly tackle the main reasons learned optimizers have not been adopted in real database systems.

Core claim

RELOAD is a robust and efficient learned query optimizer for database systems that minimizes query-level performance regressions and ensures consistent optimization behavior across executions while accelerating convergence to the plan quality of expert cost-based optimizers, demonstrated through experiments on the Join Order Benchmark, TPC-DS, and Star Schema Benchmark to achieve up to 2.4x higher robustness and 3.1x greater efficiency compared to state-of-the-art RL-based query optimization techniques.

What carries the argument

RELOAD's dual focus on minimizing query-level performance regressions for robustness and accelerating convergence to expert-level plan quality for efficiency.

If this is right

  • Reinforcement learning based query optimizers can avoid severe performance regressions on individual queries.
  • Training for learned optimizers reaches expert plan quality with substantially less time and compute.
  • Optimization behavior becomes consistent across repeated executions of the same query.
  • Learned query optimizers become easier to deploy in production database systems by addressing instability and slow training.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If the gains hold beyond the tested benchmarks, learned methods could replace or augment cost-based optimizers in more database systems.
  • The same robustness techniques might reduce risk when applying reinforcement learning to other database tuning tasks like indexing or resource allocation.
  • Production systems could adopt such optimizers with lower monitoring overhead if regressions are reliably bounded.

Load-bearing premise

The robustness and efficiency improvements measured on three standard benchmarks will hold for arbitrary real-world database workloads, schema changes, and hardware without extra tuning or unexpected drops.

What would settle it

A new benchmark workload or production trace where RELOAD produces at least one query plan whose execution time exceeds the best traditional optimizer by a larger margin than reported in the paper's experiments.

Figures

Figures reproduced from arXiv: 2604.14725 by Jaeyoung Sim, Kwanghyun Park, Seokwon Lee, Sihyun Kim, Yiwen Zhu, Yuhsing Li.

Figure 1
Figure 1. Figure 1: Challenges encountered by an RL-based query optimizer on the [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: (a) and (b) illustrate two trends of performance regression observed in [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: RELOAD integration into the RL-based Query Optimizer, with new components in dashed lines. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: PER-based Knowledge Retention in RELOAD. [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: MAML-based Knowledge Transfer in RELOAD. [PITH_FULL_IMAGE:figures/full_fig_p006_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Performance of RELOAD on test set in PostgreSQL. Shaded areas denote variation across runs. [PITH_FULL_IMAGE:figures/full_fig_p009_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Performance of RELOAD on train set in PostgreSQL. Shaded areas denote variation across runs. [PITH_FULL_IMAGE:figures/full_fig_p009_7.png] view at source ↗
Figure 9
Figure 9. Figure 9: Impact of RELOAD on SQL Server under the SSB test set. Shaded areas denoted variation across runs. limited stability on queries not encountered during training. Balsa performs competitively on JOB and SSB but underper￾forms on TPC-DS. LIMAO performs moderately but remains sensitive to query templates excluded from the training set. When combined with RELOAD, both Balsa and LIMAO exhibit faster convergence … view at source ↗
Figure 10
Figure 10. Figure 10: Micro-experiment results of RELOAD on the JOB test set. (a) evaluates the impact of different weighting strategies in knowledge retention using PER, [PITH_FULL_IMAGE:figures/full_fig_p012_10.png] view at source ↗
read the original abstract

Recent advances in query optimization have shifted from traditional rule-based and cost-based techniques towards machine learning-driven approaches. Among these, reinforcement learning (RL) has attracted significant attention due to its ability to optimize long-term performance by learning policies over query planning. However, existing RL-based query optimizers often exhibit unstable performance at the level of individual queries, including severe performance regressions, and require prolonged training to reach the plan quality of expert, cost-based optimizers. These shortcomings make learned query optimizers difficult to deploy in practice and remain a major barrier to their adoption in production database systems. To address these challenges, we present RELOAD, a robust and efficient learned query optimizer for database systems. RELOAD focuses on (i) robustness, by minimizing query-level performance regressions and ensuring consistent optimization behavior across executions, and (ii) efficiency, by accelerating convergence to expert-level plan quality. Through extensive experiments on standard benchmarks, including Join Order Benchmark, TPC-DS, and Star Schema Benchmark, RELOAD demonstrates up to 2.4x higher robustness and 3.1x greater efficiency compared to state-of-the-art RL-based query optimization techniques.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper introduces RELOAD, a reinforcement learning-based learned query optimizer for database systems. It targets two main issues in prior RL query optimizers: unstable per-query performance (including regressions) and slow convergence to the plan quality of expert cost-based optimizers. RELOAD claims to improve robustness by minimizing query-level regressions and ensuring consistent behavior, and efficiency by accelerating training. Experiments on the Join Order Benchmark, TPC-DS, and Star Schema Benchmark are reported to yield up to 2.4x higher robustness and 3.1x greater efficiency relative to state-of-the-art RL-based techniques.

Significance. If the robustness and efficiency gains are reproducible and the approach generalizes, the work could meaningfully lower barriers to deploying learned query optimizers in production systems. Addressing per-query instability and training time directly tackles practical adoption hurdles that have limited prior RL methods.

major comments (2)
  1. [Experiments] Experiments section: The reported 2.4x robustness and 3.1x efficiency gains are demonstrated only on JOB, TPC-DS, and SSB. These benchmarks share similar join structures, data scales, and query templates; no results are shown for schema evolution, data distribution shifts, or production-style ad-hoc queries. This leaves the central claim of robustness for arbitrary real-world workloads without direct support.
  2. [Methods] Methods and evaluation protocol: The abstract and reported results provide no details on experimental protocol, statistical significance testing, hyperparameter search procedure, number of runs, or controls for post-hoc selection of reported numbers. Without these, the quantitative claims cannot be fully assessed for soundness.
minor comments (2)
  1. [Abstract] The abstract would be strengthened by a concise statement of the core algorithmic modifications in RELOAD (e.g., specific changes to the RL policy or reward function) rather than remaining at a high-level description.
  2. Notation for robustness and efficiency metrics should be defined explicitly when first introduced to aid readability.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive and detailed review. We address each major comment point by point below, providing honest clarifications based on the current manuscript and indicating planned revisions where appropriate.

read point-by-point responses

  1. Referee: [Experiments] Experiments section: The reported 2.4x robustness and 3.1x efficiency gains are demonstrated only on JOB, TPC-DS, and SSB. These benchmarks share similar join structures, data scales, and query templates; no results are shown for schema evolution, data distribution shifts, or production-style ad-hoc queries. This leaves the central claim of robustness for arbitrary real-world workloads without direct support.

    Authors: We agree that the evaluation is confined to JOB, TPC-DS, and SSB, which are standard benchmarks in the query optimization literature but do not encompass schema evolution, data distribution shifts, or fully ad-hoc production queries. However, the manuscript does not assert robustness for arbitrary real-world workloads; the claims of up to 2.4x higher robustness and 3.1x greater efficiency are explicitly tied to results on these three benchmarks, which include diverse join orders, query templates, and data scales. To address the concern, we will make a partial revision by adding an explicit limitations subsection in the revised manuscript that discusses the scope of the claims, potential generalization challenges, and directions for future work on more dynamic workloads. No new experiments are added at this stage, as the current results remain valid for the evaluated settings. revision: partial

  2. Referee: [Methods] Methods and evaluation protocol: The abstract and reported results provide no details on experimental protocol, statistical significance testing, hyperparameter search procedure, number of runs, or controls for post-hoc selection of reported numbers. Without these, the quantitative claims cannot be fully assessed for soundness.

    Authors: We acknowledge that the manuscript currently provides insufficient detail on the experimental protocol, which limits full assessment of the results. In the revised version, we will expand the Experiments section (and add a dedicated subsection if needed) to include: the complete experimental protocol; statistical significance testing (reporting means, standard deviations over multiple runs, and p-values from paired t-tests); the hyperparameter search procedure and ranges; the number of independent runs (five runs with different random seeds); and controls against post-hoc selection (pre-specified metrics and reporting of all relevant outcomes). These additions will directly improve the soundness and reproducibility of the quantitative claims. revision: yes

Circularity Check

0 steps flagged

No circularity: empirical claims rest on external benchmarks

full rationale

The paper is an empirical systems contribution whose central claims (robustness and efficiency gains) are established solely by direct experimental comparison against prior RL-based optimizers on the fixed, publicly available JOB/TPC-DS/SSB workloads. No equations, parameter-fitting steps, uniqueness theorems, or ansatzes appear in the abstract or described derivation chain; the reported 2.4×/3.1× factors are measured outcomes, not quantities defined in terms of themselves or recovered by construction from the same training data. The work therefore contains no self-definitional, fitted-input-called-prediction, or self-citation-load-bearing reductions.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

The central claim rests on standard reinforcement-learning assumptions (Markov decision process formulation of query planning, reward signals based on execution time) and the representativeness of the three cited benchmarks; no new entities or ad-hoc axioms are introduced in the abstract.

pith-pipeline@v0.9.0 · 5518 in / 1173 out tokens · 57708 ms · 2026-05-10T10:05:13.194591+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

43 extracted references · 13 canonical work pages

  1. [1]

    LEON: A New Framework for ML- Aided Query Optimization,

    X. Chen, H. Chen, Z. Liang, S. Liu, J. Wang, K. Zeng, H. Su, and K. Zheng, “LEON: A New Framework for ML- Aided Query Optimization,”Proceedings of the VLDB Endowment, vol. 16, no. 9, pp. 2261–2273, May 2023. [Online]. Available: https://dl.acm.org/doi/10.14778/3598581.3598597

    work page doi:10.14778/3598581.3598597 2023

  2. [2]

    How good are query optimizers, really?

    V . Leis, A. Gubichev, A. Mirchev, P. Boncz, A. Kemper, and T. Neumann, “How good are query optimizers, really?”Proc. VLDB Endow., vol. 9, no. 3, pp. 204–215, Nov. 2015. [Online]. Available: https://dl.acm.org/doi/10.14778/2850583.2850594 JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 13

    work page doi:10.14778/2850583.2850594 2015

  3. [3]

    Marcus, P

    R. Marcus, P. Negi, H. Mao, C. Zhang, M. Alizadeh, T. Kraska, O. Papaemmanouil, and N. Tatbul, “Neo: A Learned Query Optimizer,” Proceedings of the VLDB Endowment, vol. 12, no. 11, pp. 1705–1718, Jul. 2019, arXiv:1904.03711 [cs]. [Online]. Available: http://arxiv.org/abs/1904.03711

    work page arXiv 2019

  4. [4]

    Query optimization,

    Y . E. Ioannidis, “Query optimization,”ACM Computing Surveys, vol. 28, no. 1, pp. 121–123, Mar. 1996

    1996

  5. [5]

    Iterative dynamic programming: A new class of query optimization algorithms,

    D. Kossmann and K. Stocker, “Iterative dynamic programming: A new class of query optimization algorithms,”ACM Transactions on Database Systems, vol. 25, no. 1, pp. 43–82, Mar. 2000

    2000

  6. [6]

    Model-based Reinforcement Learning: A Survey,

    T. M. Moerland, J. Broekens, A. Plaat, and C. M. Jonker, “Model-based Reinforcement Learning: A Survey,” Mar. 2022

    2022

  7. [7]

    Learn- ing to Optimize Join Queries With Deep Reinforcement Learning,

    S. Krishnan, Z. Yang, K. Goldberg, J. Hellerstein, and I. Stoica, “Learn- ing to Optimize Join Queries With Deep Reinforcement Learning,” Jan. 2019

    2019

  8. [8]

    Towards a Hands-Free Query Optimizer through Deep Learning,

    R. Marcus and O. Papaemmanouil, “Towards a Hands-Free Query Optimizer through Deep Learning,” Dec. 2018

    2018

  9. [9]

    R. S. Sutton and A. G. Barto,Reinforcement Learning: An Introduction, ser. Adaptive Computation and Machine Learning. Cambridge, Mass: MIT Press, 1998

    1998

  10. [10]

    Learning to predict by the methods of temporal differ- ences,

    R. S. Sutton, “Learning to predict by the methods of temporal differ- ences,”Machine Learning, vol. 3, no. 1, pp. 9–44, Aug. 1988

    1988

  11. [11]

    Bao: Making Learned Query Optimization Practical,

    R. Marcus, P. Negi, H. Mao, N. Tatbul, M. Alizadeh, and T. Kraska, “Bao: Making Learned Query Optimization Practical,” in Proceedings of the 2021 International Conference on Management of Data, ser. SIGMOD ’21. New York, NY , USA: Association for Computing Machinery, Jun. 2021, pp. 1275–1288. [Online]. Available: https://dl.acm.org/doi/10.1145/3448016.3452838

    work page doi:10.1145/3448016.3452838 2021

  12. [12]

    Balsa: Learning a Query Optimizer Without Expert Demonstrations,

    Z. Yang, W.-L. Chiang, S. Luan, G. Mittal, M. Luo, and I. Stoica, “Balsa: Learning a Query Optimizer Without Expert Demonstrations,” inProceedings of the 2022 International Conference on Management of Data, ser. SIGMOD ’22. New York, NY , USA: Association for Computing Machinery, Jun. 2022, pp. 931–944. [Online]. Available: https://dl.acm.org/doi/10.1145/...

    work page doi:10.1145/3514221.3517885 2022

  13. [13]

    LOGER: A Learned Optimizer Towards Generating Efficient and Robust Query Execution Plans,

    T. Chen, J. Gao, H. Chen, and Y . Tu, “LOGER: A Learned Optimizer Towards Generating Efficient and Robust Query Execution Plans,”Proceedings of the VLDB Endowment, vol. 16, no. 7, pp. 1777–1789, Mar. 2023. [Online]. Available: https://dl.acm.org/doi/10. 14778/3587136.3587150

    work page arXiv 2023

  14. [14]

    LIMAO: A Framework for Lifelong Modular Learned Query Optimization,

    Q. Zhang, S. Xie, and I. Sabek, “LIMAO: A Framework for Lifelong Modular Learned Query Optimization,” Jun. 2025

    2025

  15. [15]

    Glo: Towards generalized learned query optimization,

    T. Chen, J. Gao, Y . Tu, and M. Xu, “Glo: Towards generalized learned query optimization,” in2024 IEEE 40th International Conference on Data Engineering (ICDE). IEEE, 2024, pp. 4843–4855

    2024

  16. [16]

    Eraser: Eliminating performance regression on learned query optimizer,

    L. Weng, R. Zhu, D. Wu, B. Ding, B. Zheng, and J. Zhou, “Eraser: Eliminating performance regression on learned query optimizer,”Pro- ceedings of the VLDB Endowment, vol. 17, no. 5, pp. 926–938, 2024

    2024

  17. [17]

    Athena: An effective learning-based framework for query optimizer performance improvement,

    R. Li, Q. Li, H. Liu, R. Mao, Q. Li, and B. Tang, “Athena: An effective learning-based framework for query optimizer performance improvement,”Proceedings of the ACM on Management of Data, vol. 3, no. 3, pp. 1–24, 2025

    2025

  18. [18]

    Transfer Learning in Deep Reinforcement Learning: A Survey,

    Z. Zhu, K. Lin, A. K. Jain, and J. Zhou, “Transfer Learning in Deep Reinforcement Learning: A Survey,”IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 45, no. 11, pp. 13 344–13 362, Nov. 2023. [Online]. Available: https://ieeexplore.ieee.org/document/ 10172347/

    2023

  19. [19]

    Towards Continual Reinforcement Learning: A Review and Perspectives,

    K. Khetarpal, M. Riemer, I. Rish, and D. Precup, “Towards Continual Reinforcement Learning: A Review and Perspectives,”Journal of Arti- ficial Intelligence Research, vol. 75, pp. 1401–1476, Dec. 2022

    2022

  20. [20]

    Self-improving reactive agents based on reinforcement learn- ing, planning and teaching,

    L.-J. Lin, “Self-improving reactive agents based on reinforcement learn- ing, planning and teaching,”Machine Learning, vol. 8, no. 3, pp. 293– 321, May 1992

    1992

  21. [21]

    Prioritized Experience Replay,

    T. Schaul, J. Quan, I. Antonoglou, and D. Silver, “Prioritized Experience Replay,” Feb. 2016

    2016

  22. [22]

    Transfer Learning for Reinforcement Learning Domains: A Survey,

    M. E. Taylor and P. Stone, “Transfer Learning for Reinforcement Learning Domains: A Survey,”J. Mach. Learn. Res., vol. 10, pp. 1633– 1685, 2009

    2009

  23. [23]

    Learning to learn: Introduction and overview,

    S. Thrun and L. Pratt, “Learning to learn: Introduction and overview,” inLearning to learn. Springer, 1998, pp. 3–17

    1998

  24. [24]

    A Survey of Meta-Reinforcement Learning,

    J. Beck, R. Vuorio, E. Z. Liu, Z. Xiong, L. Zintgraf, C. Finn, and S. Whiteson, “A Survey of Meta-Reinforcement Learning,” Aug. 2024

    2024

  25. [25]

    Learning State Representations for Query Optimization with Deep Reinforcement Learning,

    J. Ortiz, M. Balazinska, J. Gehrke, and S. S. Keerthi, “Learning State Representations for Query Optimization with Deep Reinforcement Learning,” inProceedings of the Second Workshop on Data Management for End-To-End Machine Learning, ser. DEEM’18. New York, NY , USA: Association for Computing Machinery, 2018, pp. 1–4

    2018

  26. [26]

    Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks,

    C. Finn, P. Abbeel, and S. Levine, “Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks,” inProceedings of the 34th International Conference on Machine Learning. PMLR, Jul. 2017, pp. 1126–1135

    2017

  27. [27]

    Generalization of Model- Agnostic Meta-Learning Algorithms: Recurring and Unseen Tasks,

    A. Fallah, A. Mokhtari, and A. Ozdaglar, “Generalization of Model- Agnostic Meta-Learning Algorithms: Recurring and Unseen Tasks,” in Advances in Neural Information Processing Systems, vol. 34. Curran Associates, Inc., 2021, pp. 5469–5480

    2021

  28. [28]

    Rapid Learning or Feature Reuse? Towards Understanding the Effectiveness of MAML,

    A. Raghu, M. Raghu, S. Bengio, and O. Vinyals, “Rapid Learning or Feature Reuse? Towards Understanding the Effectiveness of MAML,” Feb. 2020

    2020

  29. [29]

    A Cluster Separation Measure,

    D. L. Davies and D. W. Bouldin, “A Cluster Separation Measure,”IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. PAMI- 1, no. 2, pp. 224–227, Apr. 1979

    1979

  30. [30]

    An extensive comparative study of cluster validity indices,

    O. Arbelaitz, I. Gurrutxaga, J. Muguerza, J. M. P ´erez, and I. Perona, “An extensive comparative study of cluster validity indices,”Pattern Recognition, vol. 46, no. 1, pp. 243–256, Jan. 2013

    2013

  31. [31]

    Measuring Query Complexity in SQLShare Workload,

    A. Vashistha and S. Jain, “Measuring Query Complexity in SQLShare Workload,” inProceedings of the 2019 International Conference on Management of Data, 2016

    2019

  32. [32]

    A Survey on Advancing the DBMS Query Optimizer: Cardinality Estimation, Cost Model, and Plan Enumeration,

    H. Lan, Z. Bao, and Y . Peng, “A Survey on Advancing the DBMS Query Optimizer: Cardinality Estimation, Cost Model, and Plan Enumeration,” Jan. 2021. [Online]. Available: https://arxiv.org/abs/2101.01507v1

    work page arXiv 2021

  33. [33]

    How to Distribute Data across Tasks for Meta-Learning?

    A. Cioba, M. Bromberg, Q. Wang, R. Niyogi, G. Batzolis, J. Garcia, D.-s. Shiu, and A. Bernacchia, “How to Distribute Data across Tasks for Meta-Learning?”Proceedings of the AAAI Conference on Artificial Intelligence, vol. 36, no. 6, pp. 6394–6401, Jun. 2022

    2022

  34. [34]

    FOSS: A Self-Learned Doctor for Query Optimizer,

    K. Zhong, L. Sun, T. Ji, C. Li, and H. Chen, “FOSS: A Self-Learned Doctor for Query Optimizer,” in2024 IEEE 40th International Conference on Data Engineering (ICDE), May 2024, pp. 4329–4342, arXiv:2312.06357 [cs]. [Online]. Available: http: //arxiv.org/abs/2312.06357

    work page arXiv 2024

  35. [35]

    TPC-DS, taking decision support benchmarking to the next level,

    M. Poess, B. Smith, L. Kollar, and P. Larson, “TPC-DS, taking decision support benchmarking to the next level,” inProceedings of the 2002 ACM SIGMOD International Conference on Management of Data, ser. SIGMOD ’02. New York, NY , USA: Association for Computing Machinery, 2002, pp. 582–587

    2002

  36. [36]

    The star schema bench- mark and augmented fact table indexing,

    P. O’Neil, E. O’Neil, X. Chen, and S. Revilak, “The star schema bench- mark and augmented fact table indexing,” inPerformance Evaluation and Benchmarking, R. Nambiar and M. Poess, Eds. Berlin, Heidelberg: Springer Berlin Heidelberg, 2009, pp. 237–252

    2009

  37. [37]

    Is Your Learned Query Optimizer Behaving As You Expect? A Machine Learning Perspective,

    C. Lehmann, P. Sulimov, and K. Stockinger, “Is Your Learned Query Optimizer Behaving As You Expect? A Machine Learning Perspective,” Proc. VLDB Endow., vol. 17, no. 7, pp. 1565–1577, May 2024. [Online]. Available: https://dl.acm.org/doi/10.14778/3654621.3654625

    work page doi:10.14778/3654621.3654625 2024

  38. [38]

    Why You Should Run TPC-DS: A Workload Analysis,

    M. P ¨oss, R. Nambiar, and D. Walrath, “Why You Should Run TPC-DS: A Workload Analysis,” Sep. 2007. [Online]. Available: https://www.semanticscholar.org/paper/ Why-You-Should-Run-TPC-DS%3A-A-Workload-Analysis-P% C3%B6ss-Nambiar/660aa29ca30b1e73f7a85abd97496435b76e0e8d

    2007

  39. [39]

    Variations of the star schema benchmark to test the effects of data skew on query performance,

    T. Rabl, M. Poess, H.-A. Jacobsen, P. O’Neil, and E. O’Neil, “Variations of the star schema benchmark to test the effects of data skew on query performance,” inProceedings of the 4th ACM/SPEC International Conference on Performance Engineering, ser. ICPE ’13. New York, NY , USA: Association for Computing Machinery, 2013, pp. 361–372

    2013

  40. [40]

    Learned Query Optimizer: What is New and What is Next,

    R. Zhu, L. Weng, B. Ding, and J. Zhou, “Learned Query Optimizer: What is New and What is Next,” inCompanion of the 2024 International Conference on Management of Data, ser. SIGMOD/PODS ’24. New York, NY , USA: Association for Computing Machinery, Jun. 2024, pp. 561–569. [Online]. Available: https://dl.acm.org/doi/10.1145/3626246.3654692

    work page doi:10.1145/3626246.3654692 2024

  41. [41]

    Deep Reinforcement Learning for Join Order Enumeration,

    R. Marcus and O. Papaemmanouil, “Deep Reinforcement Learning for Join Order Enumeration,” inProceedings of the First International Workshop on Exploiting Artificial Intelligence Techniques for Data Management, Jun. 2018, pp. 1–4, arXiv:1803.00055 [cs]. [Online]. Available: http://arxiv.org/abs/1803.00055

    work page arXiv 2018

  42. [42]

    Lero: A Learning-to-Rank Query Optimizer,

    R. Zhu, W. Chen, B. Ding, X. Chen, A. Pfadler, Z. Wu, and J. Zhou, “Lero: A Learning-to-Rank Query Optimizer,”Proc. VLDB Endow., vol. 16, no. 6, pp. 1466–1479, Feb. 2023. [Online]. Available: https://dl.acm.org/doi/10.14778/3583140.3583160

    work page doi:10.14778/3583140.3583160 2023

  43. [43]

    Eraser: Eliminating Performance Regression on Learned Query Optimizer,

    L. Weng, R. Zhu, D. Wu, B. Ding, B. Zheng, and J. Zhou, “Eraser: Eliminating Performance Regression on Learned Query Optimizer,” Proc. VLDB Endow., vol. 17, no. 5, pp. 926–938, May 2024. [Online]. Available: https://dl.acm.org/doi/10.14778/3641204.3641205

    work page doi:10.14778/3641204.3641205 2024