Important

ChronoLog MCP is now available. Integrate ChronoLog directly with LLMs through our new MCP server for real-time logging, event processing, and structured interactions.
Code - Documentation

ChronoLog

Distributed Shared Tiered Log Store

A distributed and tiered shared log storage ecosystem that uses physical time to distribute log entries while providing total log ordering.

Members

Illinois Tech

UChicago

Overview

ChronoLog is a distributed, tiered shared log storage system that provides scalable log storage with time-based data ordering and total log ordering guarantees. By leveraging physical time for data distribution and utilizing multiple storage tiers for elastic capacity scaling, ChronoLog eliminates the need for a central sequencer while maintaining high performance and scalability.

The system's modular, plugin-based architecture serves as a foundation for building scalable applications, including SQL-like query engines, streaming processors, log-based key-value stores, and machine learning integration modules.

Key Features

ChronoLog is built on four foundational pillars:

• Time-Structured Ingestion — Events are chunked and organized by physical time, enabling high-throughput parallel writes without a central sequencer.

• Tiered & Efficient Storage — StoryChunks flow across fast and scalable storage tiers, automatically balancing performance and capacity.

• Concurrent Access at Scale — Multi-writer, multi-reader support with zero coordination overhead, optimized for both RDMA and TCP networks.

• Modular, Extensible Serving Layer — Plugin-based architecture enables custom services to run directly on the log, supporting diverse application requirements.

Use Cases

ChronoLog's flexible architecture supports a wide range of applications:

• AI & LLM Integration — MCP server for seamless LLM integration with enterprise logging and real-time event processing.

• Machine Learning & Training — TensorFlow module for training and inference workflows using time-ordered data streams.

• SQL-like Query Engine — Query and analyze log data with SQL semantics and time-based distribution.

• Streaming Processor — Real-time event processing and analytics for monitoring, alerting, and data pipelines.

• Log-based Key-Value Store — Distributed key-value stores with strong consistency guarantees.

For more information, visit chronolog.dev.

Installation

Note

ChronoLog is designed for deployment on distributed clusters and complex environments. For comprehensive installation guides, configuration details, and deployment strategies, please refer to our detailed documentation on the Wiki.

🐳 Docker Installation

Single-Node Deployment

Pull Docker image:

docker pull gnosisrc/chronolog:latest

Run container interactively:

docker run -it --rm --name chronolog-instance gnosisrc/chronolog:latest

Deploy components:

/home/grc-iit/chronolog_repo/deploy/local_single_user_deploy.sh -d -w /home/grc-iit/chronolog_install/Release

Verify deployment:

pgrep -la chrono

For detailed setup instructions, troubleshooting, and advanced configuration, see: Single-node Docker tutorial

Multi-Node Deployment

Pull Docker image:

docker pull gnosisrc/chronolog:latest

Download deployment script:

wget https://raw.githubusercontent.com/grc-iit/ChronoLog/refs/heads/develop/CI/docker/dynamic_deploy.sh

chmod +x dynamic_deploy.sh

Run default deployment:

./dynamic_deploy.sh -n 7 -k 2 -g 2 -p 2

Verify containers:

docker ps

For detailed setup instructions, troubleshooting, and advanced configuration, see: Multi-Node Docker tutorial

Documentation

Comprehensive documentation and tutorials are available on our Wiki. The documentation covers everything from getting started to advanced configuration and deployment strategies.

📚 Documentation

Document	Description
Getting Started	Introduction and first steps with ChronoLog
Installation	Detailed installation guides and requirements
Configuration	Configuration options and settings
Deployment	Deployment strategies and best practices
Client API	API reference and usage examples
Client Examples	Code examples and use cases
Architecture	System architecture and design principles
Plugins	Plugin development and integration
Code Style Guidelines	Coding standards and conventions
Contributors Guidelines	Guidelines for contributing to ChronoLog

🎓 Tutorials

Tutorial	Description
Tutorial 1: First Steps with ChronoLog	Get started with your first ChronoLog deployment
Tutorial 2: How to run a Performance test	Learn how to benchmark and test ChronoLog performance
Tutorial 3: Running ChronoLog with Docker (single-node)	Deploy ChronoLog on a single node using Docker
Tutorial 4: Running ChronoLog with Docker (Multi-Node)	Deploy ChronoLog across multiple nodes using Docker

Collaborators

We are grateful for the collaboration and support from our research and industry partners.

Organization	Description
Argonne National Laboratory	Collaborating with the funcX team to enable event-based and real-time computing capabilities, supporting scalable execution of machine learning workloads and integration with Colmena framework for materials science applications.
University of Chicago	Working with Tom Glanzman and the Dark Energy Science Collaboration on Parsl workflow extensions for Rubin Observatory data processing, enabling extreme-scale logging and monitoring for cosmology workflows.
Parsl	Developing workflow extensions to enable extreme-scale logging and monitoring for large-scale scientific workflows, with potential impact across domains including biology, social science, and high energy physics.
Institute for Food Safety and Health (IFSH)	Exploring new scientific applications of ChronoLog in genomics and bioinformatics, with discussions helping shape development priorities while informing adjacent research communities about ChronoLog capabilities.
Lawrence Livermore National Laboratory	Working with the system scheduler team to integrate ChronoLog with Sonar and Flux job scheduler, eliminating bottlenecks in HPC resource management and telemetry data collection.
University of Wisconsin-Madison	Working with Shaowen Wang to deploy ChronoLog as a storage backend for CyberGIS, addressing growing data volume and velocity demands while refining ChronoLog features through GIS workloads.
University of Illinois at Urbana-Champaign	Collaborating on Parsl workflow extensions for the Dark Energy Science Collaboration, enabling extreme-scale logging and monitoring for Rubin Observatory data processing workflows.
DePaul University	Working with Tanu Malik to develop novel lightweight indexing mechanisms within the ChronoKeeper for efficient querying of log data by both identifier and value predicates.
ParaTools, Inc.	Exploring integration and evaluation of ChronoLog with performance monitoring tools, optimizing ChronoLog and its native plugins for application performance analysis use cases.
OmniBond Systems LLC	Working with Boyd Wilson to fine-tune the storage stack using extended attributes in OrangeFS, optimizing ChronoLog's multi-tiered distributed log store performance.

Resources

• Documentation: Visit chronolog.dev for comprehensive documentation and guides
• GitHub Repository: github.com/grc-iit/ChronoLog
• Issues & Support: Report issues or request features on GitHub Issues
• Releases: Check out the latest releases on GitHub Releases

---

Gnosis Research Center
Illinois Institute of Technology
Advancing the Future of Scalable Computing and Data-Driven Discovery

Connect with us:
🌐 Website • 🐦 X (Twitter) • 💼 LinkedIn • 📺 YouTube • ✉️ Email

Sponsored by:

National Science Foundation (NSF CSSI-2104013)