Honeycomb Helps Organizations with Cloud Applications Achieve Production Excellence
Announcing our investment in Honeycomb, the gold standard for observability software that gives enterprises better tools for improving the reliability and performance of customer-facing applications.
Companies across all industries develop and manage customer-facing applications whose reliability and performance are critical to revenue growth and customer loyalty. Keeping those applications running smoothly is an increasingly complex and expensive challenge.
Changes to the ways these applications are developed means that the model of keeping everything running smoothly with end-to-end testing is no longer effective, especially for environments that are running at high scale and continually shipping new features. Debugging gets harder as monolithic systems are replaced by distributed systems where every API call causes a cascade of related events. Traditional monitoring tools work better for monolithic systems but fall short for modern cloud-based applications.
All of this means that developers of distributed applications need to ship and test in production. Observability -- the ability for developers to understand what is happening inside their applications by observing how customers are interacting and using it -- is essential for testing in production.
Honeycomb is the pioneer of observability. They have been such effective evangelists for observability that enterprises who use their tool quickly discover they can’t live without them. One executive told us: “If a problem is really easy, you find it in a traditional monitoring tool in 30 seconds. If you don’t find it in 30 seconds, you will never find it with a traditional monitoring tool”. Honeycomb fills that gap.
Today we are pleased to announce our investment in Honeycomb, leading a new round of financing.
From Monitoring to Observability
Historically, enterprises have accepted that different monitoring solutions are required depending on what an application stores, where it is in the stack, and its use cases. From this matrix, new categories like Application Performance Monitoring (APM) and Log Analytics emerged. These tools provide infrastructure-centric and service-centric metrics and can tell you when something has gone wrong with your system.
Observability, on the other hand, is proactive, allowing devs to understand what happens inside an application from a customer perspective. Observability enables developers to address problems that they were not prepared for, the so-called “unknown unknowns”.
Understanding the Customer-Code Path
Traditional monitoring tools can be applied to distributed systems and do a decent job tracing across different services. While this helps discover, for example, which code path is slow, it’s insufficient for understanding more complex issues like whether there are latency issues with one particular customer or all of them.
This is one of the key advantages Honeycomb offers its customers. As developers know, users use applications in idiosyncratic ways and crash into edge cases all the time. As a result, they are exposed to bugs and slowdowns that are very real to them -- but get missed in the averaging out that happens across the customer base.
With Honeycomb, developers and operations teams can for the first time understand how individual users experience the app -- and then work back from those experiences to engineer fixes and improvements.
High Cardinality Is Honeycomb’s Superpower
Central to the Honeycomb product are events, a common abstraction for representing structured log lines and traces (which can be thought of as structured log lines with parent/child relationships).
Users install Honeycomb to automatically capture events and traces from all parts of the stack, including languages, cloud providers, and open source tools. Dashboards are populated and users write queries to investigate.
Unlike traditional tooling, Honeycomb customers are able to investigate the quality and performance of an individual request or a high-cardinality grouping of requests. Then customers can query on dimensions of arbitrary cardinality to understand the lifecycle of a particular event from the perspective of the user experience.
Honeycomb helps customers find problems that they did not know about, problems that negatively impact the experience of their users.
Honeycomb Customers Don’t Know How They Lived Without It
The technical advantages of Honeycomb’s observability solution were compelling enough. Then we started hearing customer after customer tell us that they can’t live without them.
We heard repeatedly that traditional monitoring tools were fine for answering “Do we have an outage and how bad is it?”. But not, “How do we go about troubleshooting it and finding where in the system it occurred?”. After adopting Honeycomb, one executive told us it was more effective than all three solutions he was using previously. And the more difficult the problem, the higher the chance he could figure it out in Honeycomb.
It’s rare to find in a young company with such a mature solution that applies to a broad set of use cases. We look forward to partnering with Honeycomb’s trailblazing founders Charity Majors and Christine Yen as they elevate observability to the de-facto standard for understanding how software runs in the real-world. We’re happy to welcome Honeycomb to the Scale family.
Originally published September 26, 2019.