• What is binjr?
• Features
• Getting started
• Trying it out
• Getting help
• Contributing
• How is it licensed?

binjr is a time series browser; it renders time series data produced by other applications as dynamically editable representations and provides advanced features to navigate the data smoothly and efficiently (drag & drop, zoom, history, detachable tabs, advanced time-range picker).

It is a standalone client application, that runs independently of the applications that produce the data; there are no server or server-side components dedicated to binjr that need to be installed on the source.

The user experience in binjr revolves around enabling users to compose a custom view by using any of the time-series exposed by the source, simply by dragging and dropping them on the view. That view then constantly evolves as users add or remove series from different sources, while navigating through it by changing the time range, the type of chart visualization and smaller aspects such as the color or transparency for each series. Users can navigate the change history for these views, using “back” and “forward” like in a web browser, and save the current state of their session at any time to a file, in order to reopen it later or to share it with someone else.

binjr also possesses the ability to visualize time series not only as charts of numeric values, but can be customized to support visualization for any data type; for instance it features out-of-the-box a source adapter for text based log files.

Log files, produced by applications to trace their lifecycle at runtime, typically contain timestamps for each event they contain; so we can think of them as time series, but with data points being textual information instead of numerical values. In practical terms, this means that a lot of the features built into binjr to compose and navigate time series visualizations can be applied to log files with great benefits.

Behind the scene, binjr uses Apache Lucene to index data from log files; meaning users can use its powerful query language to hack through vast quantities logged events. It also allows binjr to open log files of any size; unlike most text editors which will fail to load multi gigabytes-sized files as they try to fit it all in memory, binjr will happily index those and present a paginated view so that memory usage remains reasonable, while the backing index ensures that searches are fast and navigation snappy.

With these abilities, binjr aims to become the missing link between text editors and command line tools traditionally used to analyse monitoring data locally and full-blown log analytics platforms (e.g. Elastic/Logstash/Kibana stack) that centralizes logs for entire organizations. It provides many of the same powerful visualization and search features while still remaining a totally local solution (the data never needs to be pushed to the cloud - or anywhere else for that matter), and requiring no setup nor maintenance to speak of.

• binjr is not a system performance collector, nor a collector of anything else for that matter. What it provides is efficient navigation and pretty presentation for time series collected elsewhere.
• binjr is not a cloud solution. It’s not even a server based solution; it’s entirely a client application, albeit one that can get its data from remote servers. Think of it as a browser, only just for time series.
• binjr is not a live system monitoring dashboard. While you can use it to connect to live sources, its feature set is not geared toward that particular task, and there are better tools for that out there. Instead, it aims to be an investigation tool, for when you don’t necessarily know what you’re looking for beforehand and you’ll want to build and change the view of the data as you navigate through it rather than be constrained by pre-determined dashboards.

• Standalone, client-side application.
• Can connect to any number of sources, of different types, at the same time.
• Communicates though the APIs exposed by the source.
• Supports for data sources is extensible via plugins.
• Supports time-series with numeric (e.g. charts) or text (e.g. logs) values.

• Drag and drop series from any sources directly on the chart view.
• Mix series from different sources on the same view.
• Allows charts overlay: create charts with several Y axis and a shared time line.
• Highly customizable views; choose chart types, change series colours, transparency, legends, etc...
• Save you work session to a file at any time, to be reopened later or shared with someone else.

• Mouse driven zoom of both X and Y axis.
• Drag and drop composition.
• Browser-like, forward & backward navigation of zoom history.
• Advanced time-range selection widget.
• The tabs holding the chart views can be detached into separate windows.
• Charts from different tabs/windows can be synchronized to a common time line.

• Built on top of JavaFX for a modern look and cross-platform, hardware accelerated graphics.
• Three different UI themes, to better integrate with host OS and fit user preferences.

• Cross-platform: works great on Linux, macOS and Windows desktops!
• Strong performances, even under heavy load (dozens of charts with dozens of series and thousands of samples).

binjr can consume time series data provided by the following data sources:

[1]: Support for data sources not marked as 'Built-in' requires additional plugins.

There are several ways to get up and running with binjr:

The simplest way to start using binjr is to download an application bundle from the download page.

These bundles contain all the dependencies required to run the app, including a copy of the Java runtime specially crafted to only include the required components and save disk space.
They are only ~80 MiB in size and there is one for each of the supported platform: Windows, Linux and macOS.

Simply download the one for your system, unpack it and run binjr to start!

You can also build or run the application from the source code using the included Gradle wrapper.
Simply clone the repo from Github and run:

• ./gradlew build to build the JAR for the all the modules.
• ./gradlew run to build and start the application straight away.
• ./gradlew clean packageDistribution to build an application bundle for the platform on which you ran the build.

Please note that it is mandatory to run the clean task in between two executions of the packageDistribution in the same environment.

If you'd like to experience binjr's visualization capabilities but do not have a compatible data source handy, you can use the demonstration data adapter.

It is a plugin which embeds a small, stand-alone data source that you can readily browse using binjr.

1. Make sure binjr is installed on your system and make a note of the folder it is installed in.
2. Download the binjr-adapter-demo-1.x.x.zip archive from https://github.com/binjr/binjr-adapter-demo/releases/latest
3. Copy the binjr-adapter-demo-1.x.x.jar file contained in the zip file into the plugins folder of your binjr installation.
4. Start binjr (or restart it if it was runnning when you copied the plugin) and open the demo.bjr workspace contained in the zip (from the command menu, select Workspaces > Open..., or press Ctrl+O)

The documentation can be found here.

If you encounter an issue, or would like to suggest an enhancement or a new feature, you may do so here.

binjr is released under the Apache License version 2.0.

This project accepts contributions via GitHub pull requests.

By contributing to this project you agree to the Developer Certificate of Origin (DCO). This document was created by the Linux Kernel community and is a simple statement that you, as a contributor, have the legal right to make the contribution. See the DCO file for details.