Infino Save Abandoned
Infino - Next-gen Observability. Faster, Cheaper, Smarter.
Store, search, and analyze telemetry data at scale with OpenSearch.
Table of Contents
Infino Overview
What is Infino?
Infino is a next-generation telemetry plugin for OpenSearch that efficiently handles logs, metrics, and traces at scale. It transforms OpenSearch into a unified, state-of-the-art, AI-driven observability stack.
Why Infino?
Telemetry data volumes are increasing exponentially, yet there is no purpose-built search engine for telemetry. Most observability stacks are built on Lucene, ClickHouse, or Prometheus, which are powerful tools but are not built for modern telemetry data. As a result, the stacks become complex and expensive to manage. Infino's goal is to rethink observability with a smart, high-performance telemetry engine for OpenSearch.
How it works
Infino defines a set of telemetry data as a collection, which is effectively a set of automanaged indexes in OpenSearch. A collection consists internally of optimized indexes for both documents (Lucene) and telemetry (Infino) that are searchable together. Collections also have built-in AI co-pilots to do summarizations and root cause analysis.
You can access a collection just like you would a normal OpenSearch index, except you don't have to worry about managing clusters or optimizing your index as collections are automatically scaled and optimized. When you access a collection you just need to preface your index with /infino as shown in the diagram below. You can read more about accessing OpenSearch here.
Infino collections are a great option for security and observability use cases, particularly if users do not want to manage and scale OpenSearch themselves. Infino collections enable a much faster/cheaper/simpler version of OpenSearch for security and observability that accelerate root-cause analysis by combining code, tickets, telemetry, etc.
It is worth noting that a collection is represented by a mirror Lucene index prefixed with infino- which contains stats about the collection and otherwise generally behaves as a normal index in OpenSearch. The caveat is that since collections are autoscaled and automanaged, the number of shards and replicas etc. for an Infino collection will always show as 1 and several index operations like split, merge, replicate, etc. are not honored as they don’t make sense for collections.
Cluster requests work as normal but do not impact Infino Collections which are automanaged.
Features
Note that we are still very much an alpha product but we have lots on the roadmap. Our development at the moment is focused on increasing the performance of the core engine to address cost but we are starting to add features to address complexity.
Available Now
- Store logs and metrics
- OpenSearch plugin
- Basic boolean Query DSL support
- Basic OpenSearch dashboard support
Coming Soon
- DQL support
- More Query DSL coverage
- NLP
- Traces
- Observability copilot
License
Infino is distributed under the Elastic 2.0 license.
Getting started
Quickstart
Try Infino with OpenSearch using docker. Follow the steps in Infino docker compose setup to quickly deploy Infino, OpenSearch with Infino plugin and OpenSearch dashboard.
Examples
The core Infino engine has some legacy integrations which might give you a sense of Infino's capabilities. Please note that these integrations are being ported to OpenSearch as Infino is not supported as a standalone engine.
- Integration with Fluentbit - Learn how to publish telemetry to Infino using FluentBit.
- LLM monitoring with Langchain - Discover how Infino's callback in Langchain can be used for monitoring requests in real-time. Also checkout the Langchain <> Infino docs.
- LLM monitoring with OpenAI - See an example of monitoring OpenAI's services.
- Using Infino's Python client - Get started with the Python client for Infino and explore its functionalities.
Developer Resources
VS Code Dev Container
The repo has a devcontainer configuration for easily deploying the development environment in a container. Read the dev container doc to get started with the Infino dev container.
Developer Docs
Read our repo documentation here.
Contributions
Contributions are welcome and highly appreciated! To get started, check out our repo docs and the contributing guidelines. Please file an issue if you face any problems or have any questions.
Contact Us
Contact us directly if you want to discuss your use-case over virtual coffee. You can ping us on Slack or send us an email: 
.
Contributors
A big thank you to the community for making Infino possible!
Open Source Agenda Badge
