Skip to content

kmgowda/sbk-charts

BranchesTags

Repository files navigation

SBK Charts

The sbk-charts is the python application software to create the xlsx file for given single or multiple CSV file containing the SBK performance results. The generated xlsx file contains the graphs of latency percentile variations and throughput variations. The SBK framework can be used to benchmark the performance of various storage engines like RocksDB, LMDB, LevelDB, etc. and to generate the performance results in CSV format. The sbk-charts application can be used to visualize these results in a more user-friendly way.

sbk-charts uses AI to generate descriptive summaries about throughput and latency analysis

Running SBK Charts:

<SBK directory>./sbk-charts
...
(venv-sbk-charts) kmg@Mac-Studio sbk-charts % sbk-charts -h
usage: sbk-charts [-h] -i IFILES [-o OFILE] [-secs SECONDS] [-nothreads NOTHREADS] [-chat] {huggingface,lmstudio,noai,ollama} ...

SBK Charts - Storage Benchmark Visualization Tool

positional arguments:
  {huggingface,lmstudio,noai,ollama}
                        Available sub-commands

options:
  -h, --help            show this help message and exit
  -i, --ifiles IFILES   Comma-separated list of input CSV files containing benchmark results
  -o, --ofile OFILE     Output XLSX file path (default: out.xlsx)
  -secs, --seconds SECONDS
                        Timeout seconds, default : 120
  -nothreads, --nothreads NOTHREADS
                        No parallel threads, default : False
  -chat, --chat         Start interactive chat mode with AI

Please report issues at https://github.com/kmgowda/sbk-charts

Single CSV file processing

Example command with single CSV file

(venv-sbk-charts) kmg@Mac-Studio sbk-charts % sbk-charts -i ./samples/charts/sbk-file-read.csv -o ./samples/charts/sbk-file-read.xlsx

   _____   ____    _  __            _____   _    _              _____    _______    _____
  / ____| |  _ \  | |/ /           / ____| | |  | |     /\     |  __ \  |__   __|  / ____|
 | (___   | |_) | | ' /   ______  | |      | |__| |    /  \    | |__) |    | |    | (___
  \___ \  |  _ <  |  <   |______| | |      |  __  |   / /\ \   |  _  /     | |     \___ \
  ____) | | |_) | | . \           | |____  | |  | |  / ____ \  | | \ \     | |     ____) |
 |_____/  |____/  |_|\_\           \_____| |_|  |_| /_/    \_\ |_|  \_\    |_|    |_____/

Sbk Charts Version : 3.26.1.0
Input Files :  ./samples/charts/sbk-file-read.csv
Output File :  ./samples/charts/sbk-file-read.xlsx
SBK logo image found: /Users/kmg/projects/sbk-charts/images/sbk-logo.png
xlsx file : ./samples/charts/sbk-file-read.xlsx created
Time Unit : NANOSECONDS
Reading : FILE
file : ./samples/charts/sbk-file-read.xlsx updated with graphs
AI is not enabled!. you can use the subcommands [huggingface lmstudio noai ollama] to enable it.

you can see the sample fil read in csv as input file and the generated output file is file read graphs

Multiple CSV files processing

Example command with multiple CSV files

(venv-sbk-charts) kmg@Mac-Studio sbk-charts % sbk-charts -i ./samples/charts/sbk-file-read.csv,./samples/charts/sbk-rocksdb-read.csv -o ./samples/charts/sbk-file-rocksdb-read.xlsx

   _____   ____    _  __            _____   _    _              _____    _______    _____
  / ____| |  _ \  | |/ /           / ____| | |  | |     /\     |  __ \  |__   __|  / ____|
 | (___   | |_) | | ' /   ______  | |      | |__| |    /  \    | |__) |    | |    | (___
  \___ \  |  _ <  |  <   |______| | |      |  __  |   / /\ \   |  _  /     | |     \___ \
  ____) | | |_) | | . \           | |____  | |  | |  / ____ \  | | \ \     | |     ____) |
 |_____/  |____/  |_|\_\           \_____| |_|  |_| /_/    \_\ |_|  \_\    |_|    |_____/

Sbk Charts Version : 3.26.1.0
Input Files :  ./samples/charts/sbk-file-read.csv,./samples/charts/sbk-rocksdb-read.csv
Output File :  ./samples/charts/sbk-file-rocksdb-read.xlsx
SBK logo image found: /Users/kmg/projects/sbk-charts/images/sbk-logo.png
xlsx file : ./samples/charts/sbk-file-rocksdb-read.xlsx created
Time Unit : NANOSECONDS
Reading : FILE, ROCKSDB
file : ./samples/charts/sbk-file-rocksdb-read.xlsx updated with graphs
AI is not enabled!. you can use the subcommands [huggingface lmstudio noai ollama] to enable it.

you can see the sample fil read in csv and the rocksdb red in csv as input files and the generated output file is file and rocksdb read comparesion

Python Virtual Environment Setup

Prerequisites

  • Python 3.8 or higher
  • pip (Python package installer)

Setup Instructions

Setup with Python virtual environment

#create the env
python3 -m venv venv-sbk-charts

#set the env
source venv-sbk-charts/bin/activate

# install required packages
pip install -e .

# build the sbk-charts
python3 -m build

to deactivate from the venv

# deactivate the venv
deactivate

Setup with conda

# Create a new conda environment with Python 3.14 or higher
conda create -n conda-sbk-charts python=3.14 -y

# Activate the environment
conda activate conda-sbk-charts

# Install pip if not already installed
conda install pip -y

# Install the project in editable mode using pip
pip install -e .

# Build the sbk-charts package
python -m build
```ch

To deactivate

conda deactivate


## Generative AI-Powered Analysis

SBK Charts includes AI-powered analysis descriptions to provide deeper insights into your storage benchmark results.
As of today, The analysis is performed using the Hugging Face model and includes:

### Available AI Analyses

1. **Throughput Analysis**
   - Analyzes MB/s and records/s metrics
   - Identifies performance patterns and anomalies
   - Compares performance across different storage systems

2. **Latency Analysis**
   - Examines latency distributions
   - Highlights tail latency patterns
   - Provides comparative analysis between different storage configurations

3. **Total MB Analysis**
   - Analyzes total data transferred
   - Identifies throughput patterns over time
   - Compares data transfer efficiency

4. **Percentile Histogram Analysis**
   - Detailed analysis of latency percentiles
   - Identifies performance bottlenecks
   - Compares percentile distributions across storage systems

## AI Backends

SBK Charts supports multiple AI backends for analysis:

1. **LM Studio** - For local AI inference with LM Studio
2. **Ollama** - For running local LLMs through the Ollama API
3. **Hugging Face** - For cloud-based AI analysis (default)

### LM Studio Setup

1. Install [LM Studio](https://lmstudio.ai/)
2. Download and host a suitable model (e.g., Mistral 7B, Llama 3.1)
3. Start the LM Studio server

Example usage:
```bash
sbk-charts -i input.csv -o output.xlsx lmstudio

Ollama Setup

  1. Install Ollama
  2. Pull required models: 
    ollama pull llama3.1

Example usage:

sbk-charts -i input.csv -o output.xlsx ollama

For more details, see the documentation in custom AI models

Interactive Chat Mode

SBK Charts provides an interactive chat mode that allows you to query your CSV data using AI-powered analysis with a RAG (Retrieval-Augmented Generation) pipeline.

Chat Mode Features

  • Natural Language Queries: Ask questions about your storage benchmark data in plain English
  • RAG Pipeline: Uses advanced retrieval to provide context-aware answers based on your CSV data
  • Multiline Support: Type complex queries across multiple lines
  • Interactive Interface: Real-time conversation with AI about your data
  • Data Context: AI has access to all storage statistics from your benchmark results

Using Chat Mode

To start an interactive chat session:

sbk-charts -i input.csv -o output.xlsx -chat huggingface

Or with other AI backends:

# Using Ollama
sbk-charts -i input.csv -o output.xlsx -chat ollama

# Using LM Studio
sbk-charts -i input.csv -o output.xlsx -chat lmstudio

Chat Interface

Once in chat mode:

  1. Type your query and press Enter twice to submit
  2. Multiline queries are supported - just press Enter twice when done typing
  3. Press Ctrl+D to exit chat mode
  4. AI responses appear after processing with status indicators

Example Queries

You can ask questions like:

  • "Which storage system performs better for read operations?"
  • "Compare the throughput between FILE and ROCKSDB storage systems"
  • "What is the best storage system for high IOPS workloads?"
  • "Show me the latency analysis for all storage systems"
  • "Which storage has the lowest tail latency?"

Chat Mode Workflow

  1. Initialization: The chat mode initializes a RAG pipeline with your storage statistics
  2. Context Loading: All performance data from your CSV files is made available to the AI
  3. Interactive Loop: Ask questions and receive AI-powered insights
  4. Smart Responses: AI uses the RAG pipeline to provide accurate, context-aware answers

Technical Details

  • RAG Pipeline: Uses Simple RAG for maximum compatibility without external dependencies
  • Data Ingestion: Automatically processes storage statistics from all worksheets
  • Threaded Processing: AI responses run in separate threads for non-blocking interaction
  • Error Handling: Graceful handling of AI backend errors and timeouts

Contributing

We welcome and appreciate contributions from the open-source community! Whether you're interested in improving the code, enhancing documentation, or adding new AI backend models, your contributions help make SBK Charts better for everyone.

How to Contribute

  1. Fork the repository and create your feature branch (git checkout -b feature/amazing-feature)
  2. Commit your changes with clear, descriptive messages
  3. Push to the branch (git push origin feature/amazing-feature)
  4. Open a Pull Request with a clear description of your changes

Areas Needing Contributions

1. Code Improvements

  • Performance optimizations
  • Bug fixes
  • New features and enhancements
  • Test coverage improvements

2. Documentation

  • Improve existing documentation
  • Add usage examples
  • Create tutorials or guides
  • Translate documentation to other languages

3. AI Model Integrations

We're particularly interested in expanding our AI capabilities. You can help by:

  • Adding support for new AI providers (e.g., OpenAI, Anthropic, local LLMs)
  • Improving prompt engineering for better analysis
  • Adding new types of performance analysis
  • Supporting more benchmark result formats

Setting Up for Development

Set up the development environment as described in the Python Virtual Environment Setup section

Code Style

  • Follow PEP 8 for Python code
  • Use type hints for better code clarity
  • Write docstrings for all public functions and classes
  • Keep commits atomic and focused

Reporting Issues

Found a bug or have a feature request? Please open an issue on our GitHub Issues page with:

  • A clear description of the issue
  • Steps to reproduce
  • Expected vs actual behavior
  • Any relevant screenshots or logs

License

By contributing, you agree that your contributions will be licensed under the Apache License 2.0.

Release format

Release Number : .<2 digits year>.<2 digits month>. Example: 1.25.6.0 Major number - 1 Year - 25 (2025) Mont - 6 (Jun) Minor Number - 0

About

Charts for the SBK performance results and AI based analysis

Topics

charts ai graphs pandas python3 performance-analysis observability sbk generative-ai

Resources

Readme

License

Apache-2.0 license
Activity

Stars

0 stars

Watchers

0 watching

Forks

0 forks
Report repository

Releases 6

sbk-charts release version 3.26.2.1 Latest
Feb 15, 2026
+ 5 releases

Packages

 
 
 

Contributors

Languages