MétaCan
Menu
Back to cohort
Record W4414266787 · doi:10.14778/3750601.3750638

SQL:Trek Automated Index Design at Airbnb

2025· article· en· W4414266787 on OpenAlexaff
Sam Lightstone, Ping Wang

Bibliographic record

VenueProceedings of the VLDB Endowment · 2025
Typearticle
Languageen
FieldComputer Science
TopicData Management and Algorithms
Canadian institutionsAir Canada
Fundersnot available
KeywordsIndex (typography)ScalabilityIndex selectionFalse positive paradoxSelection (genetic algorithm)Database indexCompilerRelational database

Abstract

fetched live from OpenAlex

Automating index design has been an active area of research for decades due to the significant impact that indexes have on query performance and database efficiency. Existing approaches range from brute-force search to cost-based optimizations and, more recently, machine learning techniques. However, many suffer from high computational costs, reliance on inaccurate cost models, or the need for deep integration with database internals. We introduce SQL:Trek, a time-efficient tool for automated index design that operates entirely as an external utility. SQL:Trek leverages query compiler cost models to identify effective indexes while mitigating false positives through execution on a lightweight simulation database. This approach enables fast, iterative index selection without modifying database internals, making it broadly applicable across relational databases, including most MySQL ® and PostgreSQL ® derivative databases. Our evaluation demonstrates that SQL:Trek delivers significant query performance improvements while keeping index selection computationally efficient, with most workloads analyzed in under five minutes. Unlike many cost-based what-if analysis methods, SQL:Trek significantly improved performance of many production workloads while avoiding the majority of detrimental index recommendations caused by optimizer misestimates. These results highlight SQL:Trek as a practical, scalable solution for automated index tuning in modern database environments.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame distilled prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.

metaresearch head score (Codex)0.000
metaresearch head score (Gemma)0.000
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Not applicable · Consensus signal: none
GenreCandidate signal: Empirical · Consensus signal: none
Teacher disagreement score0.856
Threshold uncertainty score0.437

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0000.000
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0000.001
Science and technology studies0.0000.000
Scholarly communication0.0000.000
Open science0.0020.002
Research integrity0.0000.000
Insufficient payload (model declined to judge)0.0000.000

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.011
GPT teacher head0.230
Teacher spread0.219 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one teacher head, not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designNot applicable
Domainnot available
GenreEmpirical

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations0
Published2025
Admission routes1
Has abstractyes

Explore more

Same venueProceedings of the VLDB EndowmentSame topicData Management and AlgorithmsFrench-language works237,207