Operations | Monitoring | ITSM | DevOps | Cloud

In just a few short years, OpenTelemetry has become the second largest CNCF project behind Kubernetes and is well on its way to becoming an industry standard for collecting and exporting telemetry data. And with KubeCon + CloudNativeCon Europe 2025 just around the corner, there’s no one better to talk to about the state of OpenTelemetry than Ted Young. Ted is the co-founder of OpenTelemetry and serves on the OpenTelemetry Governance Committee.

CloudWatch metrics can be a very useful source of information for a number of AWS services that don’t produce telemetry as well as instrumented code. There are also a number of useful metrics for non-web-request based functions, like metrics on concurrent database requests. We use them at Honeycomb to get statistics on load balancers and RDS instances. The Amazon Data Firehose is able to export directly to Honeycomb as well, which makes getting the data into Honeycomb straightforward.

When you're running a Java application, the JVM is doing a ton of work behind the scenes but unless you're actively collecting its internal metrics, you're essentially flying blind. Fortunately, the JMX Prometheus Receiver paired with the JMX Java Exporter Agent offers one of the simplest and most effective ways to expose JVM performance data.

When you're running a Java application, the JVM is doing a ton of work behind the scenes but unless you're monitoring those internals, it's hard to know how your app is really performing. JVM metrics give you a window into the heart of the runtime: how much memory you're using, how often garbage collection is kicking in, how many threads are active, and where potential bottlenecks might be hiding.

If you’ve ever found yourself sifting through logs, metrics, and traces without a clear answer to why your app crashed at 2 AM, you’re not alone. Troubleshooting without the right tools can feel like chasing shadows. That’s where the right OpenTelemetry backend makes all the difference—bringing everything together and turning scattered data into a clear picture.

In the microservices world, tracking down performance issues feels like solving a mystery with pieces scattered across dozens of systems. When users report slowness, your team needs answers fast—not hours of guesswork. Distributed tracing is emerged as the solution, but implementing it effectively requires more than just understanding the basics. This guide takes you beyond the fundamentals to show you how DevOps teams and SREs can build truly effective tracing strategies.

Ever felt like you're missing crucial insights into your Java applications? The OpenTelemetry Java Agent changes that game completely. This comprehensive guide takes you beyond the basics, showing you not just how to implement it, but how to master it for maximum observability.

Debugging software, whether during development or incident investigation, often begins with a manual and error-prone process. Developers typically scatter logs and snapshots across the codebase, allowing them to trigger multiple times. They then inspect the outputs and sift through the results to identify those relevant to the issue under investigation. Developers tend to group results that stem from the same user request or transaction.