¶ Description
TPCH benchmark raw tables.
¶ Parameters
Parameters:
- Benchmark Select the type of benchmark data to generate:
• TPC-H
• TPC-H - Scale Factor (For TPC-H) Defines the data volume in GB (e.g., 1 GB)
- Table(s) to generate Select specific table(s) to generate from the benchmark (e.g., Customer, Orders, etc.)
- Optimize tables for processing Toggle to optimize table layout for downstream processing
- Reference for dates in Elapsed Time format Set a reference date (in
yyyyMMdd hh:mm:ssformat) to compute elapsed times - Conversion table to generate (For TPC-H conversion tables) Choose specific conversion table (e.g., C_MKT_SEGMENT)
- First Record Number (For Sort benchmark) Starting record number (default: 0)
- Number of Records to generate (For Sort benchmark) Specify how many records to generate (default: 1,000,000)
-Generate the table for the Sort Benchmark Confirms generating the Sort benchmark table
¶ About
Generates TPCH benchmark raw tables and other benchmark datasets for testing and performance evaluations.
This action supports the following benchmark types:
-
TPC-H
-
TPC-H conversion tables
-
Sort benchmark
It is typically used for: -
Simulating large datasets for pipeline testing
-
Benchmarking sorting, aggregation, and ETL performance
-
Generating reproducible test cases for development or documentation
Example Workflow:
- Generate synthetic benchmark data with GenBenchmarkData.
- Sort the generated data using the Sort action.
- Output sorted data to the next processing step.
¶ Use Cases
- Benchmarking ETL tools with large TPC-H datasets.
- Testing database performance with synthetic customer/order data.
- Generating Sort Benchmark datasets for stress testing.
- Creating sample datasets for pipeline documentation.
Notes
- The output structure and data size depend on the selected benchmark and parameters.
- Large TPC-H datasets may require significant disk space and memory.
- Sort Benchmark generates purely numeric data for sorting tests.
- Execution time and memory usage may vary with data volume.
¶ Benchmark Types Overview
¶ 1. TPC-H Benchmark
Generates TPC-H-compliant datasets with a configurable scale.
Example Configuration:
- Benchmark: TPC-H
- Scale Factor: 1 GB
- Table: Customer
- Optimize Tables: Enabled
Example Output:
| C_CUSTKEY | C_NAME | C_ADDRESS | C_NATIONKEY | C_PHONE |
|---|---|---|---|---|
| 1 | Customer#000000001 | Sample Address | 15 | 25-989-741-2988 |
| ... | ... | ... | ... | ... |
¶ 2. TPC-H Conversion Tables
Generates conversion datasets for TPC-H benchmarks.
Example Configuration:
- Benchmark: TPC-H conversion tables
- Conversion Table: P_BRAND
Example Output:
| P_BRAND | P_BRAND_INDEX |
|---|---|
| Brand#11 | 11 |
| Brand#12 | 12 |
¶ 3. Sort Benchmark
Generates numeric datasets for sorting performance tests.
Example Configuration:
- Benchmark: Sort benchmark
- First Record Number: 0
- Number of Records to Generate: 1,000,000
Example Output:
| ID_1 | ID_2 | ID_3 | ID_4 |
|---|---|---|---|
| 1248423239 | 1918990674 | 1230005512 | 6325099 |
| ... | ... | ... | ... |
