HiveMQ announced the availability of the HiveMQ Distributed Tracing Extension, a new feature that makes it possible to trace and debug MQTT data streams from device to cloud and back, per its maker. Complete IoT observability requires insight into three pillars: metrics, traces and logs. HiveMQ has added distributed tracing to help organizations achieve end-to-end observability and make their IoT applications more performant and resilient.
Distributed Tracing is a way to trace events and achieve a high-level overview of a message’s journey through multiple, complex systems. With the Distributed Tracing Extension, HiveMQ is the first MQTT broker to add OpenTelemetry support to provide complete transparency for every publish message that uses the HiveMQ MQTT broker, they claim. OpenTelemetry is an open standard for instrumentation that allows for interoperability across all services so organizations can achieve visibility over their entire system.
“We’re the first MQTT broker to enable true IoT observability so customers can trace MQTT data and gather diagnostic information in real-time rather than after the fact,” said Christian Götz, CEO and co-founder of HiveMQ. “IoT observability is key as it allows customers to quickly identify latency bottlenecks or reasons for failure in critical transactions and decrease the time spent resolving these issues.”
HiveMQ offers integration into a broad range of application-performance monitoring (APM) tools such as Datadog, Dynatrace and Honeycomb, or open-source alternatives like Grafana Tempo. APM tools are being adopted rapidly, but when used alone they typically have a blind spot around the MQTT data that leads to poor observability of applications. With the distributed-tracing extension, HiveMQ has solved that problem to unlock more value from expensive APM investments and shorten the time required to discover and resolve issues, HiveMQ stated.
“In a complex architecture, customers often don’t know where to start when they experience a problem,” added Götz. “Say opening the car door with a mobile application is taking five to 10 seconds instead of one second. A detailed look at where the message request traveled and how long it took at each step makes it easy to identify the root cause of latency so it can be fixed.”
