File size: 4,486 Bytes
d9db267 05338bb | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 | ---
license: cc-by-nc-4.0
task_categories:
- text-generation
- question-answering
language:
- en
tags:
- sql
- database
- query-optimization
- explain-analyze
- postgresql
- mysql
- clickhouse
- cockroachdb
- dba
- fine-tuning
pretty_name: "Enterprise Database Query Plan Optimization & Remediation Traces"
size_categories:
- 100<n<1K
---
# Enterprise Database Query Plan Optimization & Remediation Traces (Community Preview)
<p align="center">
<img src="dataset_product_cover_1024.png" alt="Dataset Banner" width="700"/>
</p>
## Overview
High-concurrency database workloads frequently degrade due to optimizer plan misestimations, unindexed correlated joins, merge-sort memory spills, and non-sargable expressions.
This repository hosts the **Community Preview (250 verified traces)** of an architectural fine-tuning dataset spanning **PostgreSQL 16, MySQL 8.0, ClickHouse 24, CockroachDB, MariaDB, and SQLite**. Each record pairs an authentic production query regression and raw execution plan (`EXPLAIN ANALYZE`) with chain-of-thought architectural diagnosis, actionable remediation DDL, and dialect-compliant optimized queries.
---
## Dataset Splits & Commercial Licensing
| Split | Records | Format | Primary Use Case | License |
| :--- | :--- | :--- | :--- | :--- |
| **Community Evaluation** (This Repo) | 250 | Parquet | Benchmarking, zero-shot eval, research | **CC BY-NC 4.0** |
| **Enterprise Commercial Corpus** | 2,500 | Parquet + JSONL | Production fine-tuning (SFT/DPO), commercial DBA agents | **Commercial Enterprise License** |
### Complete Enterprise Corpus (2,500 Records)
For production model deployment, proprietary agent training, or private RAG pipelines without non-commercial restrictions:
👉 **[Acquire Enterprise Commercial License & Dataset via Dodo Payments](https://checkout.dodopayments.com/buy/pdt_0NqD7nyGBW7fVDMd1Fm2l)**
---
## Data Structure & Schema
```json
{
"database_engine": "string",
"slow_query": "string",
"execution_plan": "string",
"diagnostic_thought": "string",
"remediation_ddl": "string",
"optimized_query": "string"
}
```
### Field Definitions
| Field | Data Type | Description |
| :--- | :--- | :--- |
| `database_engine` | `string` | Target engine dialect & exact version (e.g., `PostgreSQL 16.2`, `ClickHouse 24.3`). |
| `slow_query` | `string` | Realistic unoptimized SQL query triggering runtime performance degradation. |
| `execution_plan` | `string` | Raw optimizer output (`EXPLAIN (ANALYZE, BUFFERS)` or JSON plan) exposing node bottlenecks. |
| `diagnostic_thought` | `string` | Step-by-step reasoning diagnosing node cost, row estimate deltas, and I/O buffer behavior. |
| `remediation_ddl` | `string` | Executable DDL statements (`CREATE INDEX CONCURRENTLY`, table parameter tuning). |
| `optimized_query` | `string` | Performant rewritten query strictly adhering to engine dialect grammar. |
---
## Quickstart (Hugging Face Datasets)
```python
from datasets import load_dataset
# Load Community Preview
dataset = load_dataset("ayushjoshi2012/sql-optimization-diagnostic-traces", split="train")
# Inspect a single diagnostic trace
sample = dataset[0]
print(f"Engine: {sample['database_engine']}")
print(f"Slow Query:\n{sample['slow_query']}")
print(f"Diagnosis:\n{sample['diagnostic_thought']}")
print(f"Remediation DDL:\n{sample['remediation_ddl']}")
print(f"Rewritten Query:\n{sample['optimized_query']}")
```
---
## Core Domain Taxonomy
Traces are generated across representative enterprise schema designs:
* **Fintech & Ledger Core:** Double-entry ledger scans, implicit UUID-to-integer type coercion, balance aggregations.
* **IoT & Telematics:** Granule skip-index mismatches, partition decompression stalls, time-bucket groupings.
* **B2B SaaS & Security:** High-offset tenant audit logs, soft-delete booleans, unindexed RBAC joins.
* **Healthcare & EHR:** Unindexed composite foreign keys, leading-wildcard text filters, outdated optimizer statistics.
* **Freight & Global Logistics:** Disk-spilling merge joins on departure timestamps, Cartesian row multiplication.
* **AdTech & Gaming Telemetry:** Unindexed streaming click joins, unbounded window function sorts, leaderboard lock contention.
---
## Citation & Contact
For inquiries regarding custom synthetic domain generation, enterprise site licenses, or data additions, please commentinitiate a new discussion. |