Honeycomb is an event-driven observability platform for debugging systems, application code, and databases. Honeycomb ingests structured data and performs read-time query aggregation to support speedy analytics over ultra-rich datasets without indexes or schemas to maintain, and a fast, interactive interface.
Honeycomb is designed to help your team answer unpredictable new questions — quickly, accurately, and painlessly. It provides real-time, interactive introspection for your data at a scale that would drown or bankrupt other apps.
Honeycomb is built for collaboration, so everyone on a team can explore the system without needing root access or years of institutional knowledge.
Try our Events Demo
In this scenario we'll walk you through using observability to diagnose this specific outage. In 16 steps (~15 minutes) you'll go from "being paged" to breaking down the issue by high-cardinality events as you debug the outage.
Try our Tracing Demo
In this scenario we'll walk you through using observability (via tracing) to diagnose reported API slowness from a handful of users. In 13 steps (~15 minutes) you'll go from alert to the individual request that triggered the issue.
Incident investigation (and response) - Diagnose issues quickly and correctly by showing the effects of a single user, event, or transaction (or combination thereof) on the system.
Performance analysis - Investigate system performance by any axis to improve overall speeds. Isolate signal from noise and dive deep into tracking down performance problems.
Ongoing development - Using Honeycomb hand in hand with normal development helps teams really understand how code is running—not just on dev machines but in production as well.
“With Honeycomb, we were able to get a breakdown of message matching duration by customer and saw that one customer was a huge outlier there, accounting for 55% of a server time spent doing that operation-and over 10% of overall server CPU usage for the fleet!”
– Jamie Osler, Product Engineer
Debug your systems in real-time, in production. Honeycomb enables teams to diagnose and resolve the idiosyncratic issues that can only be seen in production—immediately as they happen.
See beyond dashboards. Honeycomb’s event-centered model lets software owners keep the context about each unit of work that is lost when pre-aggregating many disconnected metrics.
Elevate your engineering team. Honeycomb is built for modern engineering teams to collect and share knowledge, mine query histories, and peek over the shoulders of experts.
“If we did not have Honeycomb it may have taken an additional 30 minutes or so to determine what the issue was by poking around in ELK or Cloudwatch logs. Honeycomb’s visualization and honeymarkers made it obvious that the issue was related to a recent release. On top of that, Honeycomb allowed us to determine the affected users and pass that information on to our support team.””
– Michael Garski, Director of Engineering
FLEXIBLE, BLAZING FAST QUERY ENGINE
Validate hypotheses, faster -- Honeycomb's friendly query builder and speedy storage layer get you answers to your questions — fast. Every Honeycomb account has a Fast Query Window that provides real-time querying for recent data. You choose what data, and how much of it, gets the high-performance treatment.
Explore new unknowns -- The world is getting more complex, and it's becoming impossible to predict how your systems might misbehave. Store more context with each event, then query over just the fields you need, flexibly. Don't get stuck having to predict the future.
Get down to the raw data -- Honeycomb's high-speed analytics always preserve access to the raw events underneath a graph. Interesting spike? View “Raw Data” to find out what really happened around the events in question.
Trigger on high-cardinality fields -- Combine Triggers with high-cardinality queries to be notified if any interesting traffic segment (e.g. individual users, customers, Kafka partitions, etc) exceeds some defined threshold.
BUILT FOR MODERN ENGINEERING TEAMS
Keep useful work around, forever -- In Honeycomb, a permalink to a query result is just what it sounds like: a permanent link to useful graphs, even after the underlying data has aged out. Finally, be free to paste URLs in docs and post-mortems and trust that your work will stick around.
Backtrack easily with history -- Jump back to queries run in the past — Honeycomb keeps a history of every query you run (and the queries your teammates run!) so you can invalidate hypotheses and easily hit “undo.”
Capture sets of useful queries for later -- Bookmark and organize queries into a shared brain of important entry points into a dataset. Debugging is a team sport — even if the team is “past you, present you, and future you,” and Honeycomb Boards provide great jumping-off points for deeper investigation.
Discuss queries in your existing channels -- Honeycomb's Slack integration understands Honeycomb permalinks and unfurls graphs right in line with your ongoing conversations. Tag teammates via our “Share to Slack” UI, then hit “Send” to push updates and notes back to where your team communications already live.
HIGH PERFORMANCE DISTRIBUTED COLUMNAR STORE
Aggregate on the read path -- Honeycomb stores raw events and crunches through them at read time, offering the flexibility of grepping through logs with the speed of metrics tools. Our “instant metrics from structured logs” lets you find out about what's important now, not just what you predicted months ago.
Store sparse, wide events -- Context is king: so send the context you care about, when you care about it — Honeycomb thrives on data with flexible schemas. Honeycomb is your data tool that just “does the right thing.” Send the metadata you might need tomorrow, without worrying about hurting query performance today.
Optimize for speed -- Honeycomb takes advantage of a column-oriented architecture, solid state drives, and industry-standard approximation algorithms to execute your most complex queries, faster.