binjr v3 beta announcement¶
Version 3 for binjr is getting ready for release
This is a major version change, justified by some fairly notable - and breaking - changes in both the plugin API and the serialization format.
So we decided to first run a beta phase and gather feedback from the community, to make sure we’re on the right track!
These breaking changes were necessary to accommodate a new core feature; the ability to visualize time series not only as charts of numeric values, but any other type of visualization for any data type.
The first use of this new ability is the addition of a data source adapter for 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.
For instance, in the same way binjr lets you mix series from different sources on the same chart, it allows for viewing events from different files on the same page, with all events being sorted based on their timestamp.
Furthermore, binjr already lets you synchronize the time line for two or more separate tabs holding charts; you can now also link a log event tab in the same way. This means that the view for the logged event will be automatically refined as you zoom in or out on a linked chart (and vice versa).
How does it work?¶
Behind the scene, binjr uses Apache Lucene to index data from log files; which allows to efficiently sort, filter and search the meaning users can use its powerful query language to hack through vast quantities logged events,
This 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 to guarantee memory usage remains reasonable, while the backing index ensures navigating and searching is lighting fast.
With these new abilities, the goal is for binjr 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.
Tell us what you think!¶
Please give it a try and let us know what you think; what’s good and what’s not? What’s missing? What’s broken? What could be improved?
You’ll find the (Work-In-Progress) documentation for the new features here.
If you run into a bug, please open an issue here.