Honeycomb changelog honeycomb.io

The results table for heatmaps has been empty for too long. We've added a sparkline to give you a rough sense of the distribution for that particular group.

Notice that the sparkline is a histogram slice across the group: the left side corresponds to the low values from the heatmap, while the right side corresponds to the high values.

Thanks to Sophie Alpert, of Humu, for the suggestion!

Another improvement to derived columns! If you create a derived column in the Query Builder, it now immediately shows up in the details tab within the right sidebar:

You can now use a derived column for the Error, Route, User and Status fields!

To use a derived column in your schema, click "Configure", and select the derived column from the list of fields. Read more about configuring your schema on our docs page.

We've improved the alerting behavior for triggers. Moving forward, if new groups surpass the defined threshold while the trigger is in a Triggered state, Honeycomb will send an additional alert and include the set of triggered groups in the notification.

You can read more about triggers and alerting in the docs.

We've made some changes to your Query Builder:

Please visit the docs for more information: https://docs.honeycomb.io/working-with-your-data/queries/

We've added a section at the bottom of Home that shows your Recent Events or Recent Traces (if your dataset has trace events).

Now you can see your 5 most recent events or your 5 most recently completed traces.

By clicking on the expand icon on the left of the event, you can view the raw JSON of the event.

You can click on the go to trace button to jump directly to that trace.

This information refreshes every minute, and is not controlled by the top time selector, as we want to make sure you're seeing the most recent events and traces.

You can read more about these tables in our docs page.

We have updated the libhoney-rb Ruby SDK to use our batch endpoint for submitting events!

Previous versions of the SDK used a persistent HTTP connection, but still required one HTTP request per event. Using the batch endpoint allows the SDK to make fewer HTTP requests when sending events, and should provide high-volume customers with significant performance improvements.

The new version (1.13.0) can be found here: https://github.com/honeycombio/libhoney-rb/compare/v1.12.1...v1.13.0

Ruby beeline users and Ruby SDK users can both get the latest version of the gem by running bundle update libhoney

For more information, visit: https://docs.honeycomb.io/api/events/#batched-events

When we launched Home a few weeks ago, you clamored for us to add support for derived columns, especially for the Error field.

We listened and you can now configure your dataset to use a derived column in the Error field.

To configure this, first create a derived column in your dataset settings page. Then, click on the Honeycomb logo to navigate to Home, then click Configure. From there, you can choose your new derived column from the drop-down menu beside the Error name.

We calculate the Error Rate in the graph on home by calculating the average of a custom derived column. This looks like IF(BOOL(<Your-derived-column-here>), 100, 0).

You can read more about derived columns on our docs page.

Over the last few weeks, a number of cool new things have shown up in tracing. We've summarized them in this blog post: Investigating Timeouts with Tracing.

In it, we cover all of these new features:

• Direct links to traces
• Searching traces for specific values
• A minigraph to identify how unusual a specific span is
• Starting new queries from traces.

You can now use either Boolean or String type fields in the Error breakdown on Home!

If your dataset has an Error field that receives error messages as the value, the Error Rate will be calculated based on that field, and you will also be able to breakdown your data based on that field.