benchmark-runner

1.0.986 verified Fri May 01 auth: no python

A CLI tool for running and managing performance benchmarks across multiple platforms including HammerDB, VDbench, and fio. Current version 1.0.986, requires Python >=3.10, released under Red Hat Performance. Active development with frequent releases.

pip install benchmark-runner

Common errors

error ModuleNotFoundError: No module named 'benchmark_runner' ↓

cause The library might not be installed or the Python environment is incorrect.

fix

Run 'pip install benchmark-runner' and ensure you are using the correct Python environment (Python >=3.10).

error benchmark-runner: command not found ↓

cause The pip install location is not in PATH, or the package was not installed globally.

fix

Use 'python -m benchmark_runner' or add the pip scripts directory to your PATH.

error ValueError: Unsupported benchmark type 'xyz' ↓

cause The benchmark type specified is not recognized. Supported types: hammerdb, vdbench, fio, etc.

fix

Check the benchmark type string (e.g., 'fio' instead of 'io') and consult the documentation for supported types.

Warnings

gotcha The library is primarily a CLI tool; Python import usage may not be fully documented. Use the command-line interface for reliable operation. ↓

fix Prefer running 'benchmark-runner <command>' from terminal instead of importing in Python.

breaking Version 1.0.894 changed default database versions for HammerDB (MSSQL 2019→2022, PostgreSQL 10→13, MariaDB 10.3→10.5) and VDbench base OS (CentOS Stream 8→9). Existing configurations relying on old defaults may break. ↓

fix Update benchmark configurations to specify exact versions if needed, or adjust to new defaults.

deprecated Python 3.10 is the minimum supported version. Older Python versions (3.9 and below) are no longer supported as of recent releases. ↓

fix Upgrade to Python 3.10 or higher.

Imports

benchmark_runner
```
import benchmark_runner
```

Quickstart

Basic usage of benchmark-runner library to execute a fio read benchmark.

import benchmark_runner

# Run a benchmark (example with fio)
benchmark_runner.run(
    benchmark_type='fio',
    workload='read',
    output_dir='/tmp/results',
    metadata={'user': 'test'}
)