The Future of Open Telemetry

OpenTelemetry Takes Centre Stage as Cloud Native Observability’s Backbone

At KubeCon Europe 2025 in London, one message echoed clearly throughout the observability community: OpenTelemetry (OTel) is no longer a peripheral initiative, it has become the backbone of the modern observability stack. Whether it’s container runtimes, service meshes, managed platforms or self-hosted deployments, OpenTelemetry has embedded itself into the core of the cloud native ecosystem.

This is more than just widespread adoption, it represents consolidation. OpenTelemetry is fast becoming the de facto standard layer for telemetry in cloud native environments.

OpenTelemetry in Action: Insights from the KubeCon Stage

Many sessions this year showcased just how deeply OTel is now embedded in the cloud native stack:

In Jaeger V2: OpenTelemetry at the Core of Modern Distributed Tracing, Jonah Kowall demonstrated how Jaeger V2 has adopted OpenTelemetry as its foundational telemetry framework. This shift to vendor-neutral instrumentation allows Jaeger to natively ingest OpenTelemetry Protocol (OTLP), leverage enhanced semantic conventions, and delegate data collection, batching and transformation to the OpenTelemetry Collector.

In The State of Prometheus and OpenTelemetry Interoperability, Arthur Silva Sens and Juraj Michálek explored the integration challenges between these two CNCF projects. They outlined the fundamental differences, most notably, Prometheus’s pull-based model compared with OpenTelemetry’s push-based approach. The OpenTelemetry-Prometheus SIG is actively working to close these gaps, with developments including UTF-8 character support in Prometheus and improvements to OpenTelemetry’s Prometheus exporters. The session showcased ongoing efforts to enable seamless interoperability between Prometheus and OpenTelemetry.

In Expanding eBPF’s Reach: From Batteries-Included Auto-Instrumentation To E2E Observability Pipelines, Dom Delnano explored how eBPF enables dynamic, low-overhead instrumentation, particularly suited to cloud native systems. Projects such as Pixie and Inspektor Gadget were highlighted for their flexible observability capabilities. He also introduced the re-architecting of Pixie’s data collector into a universal agent, compatible with popular observability pipelines like Fluentbit, facilitating unified data processing across heterogeneous environments.

In How To Adopt OpenTelemetry in an Enterprise Where Incumbent Vendor Tools Reign Supreme, Chris Weldon shared the experience of Wolters Kluwer as they transitioned from a fragmented tooling landscape to an OpenTelemetry-based observability approach. This change aimed to improve high mean-time-to-resolution (MTTR) challenges by standardising telemetry collection and reducing tool sprawl. Weldon discussed the trade-offs in integrating OpenTelemetry within existing workflows, particularly around bespoke (“snowflake”) data requirements. He advocated for a platform-as-a-product mindset, positioning observability as a shared organisational service to drive adoption and enhance operational maturity.​

Observability Day Europe, co-located with KubeCon, also featured multiple sessions showing how OTel is now being used to unify telemetry from container orchestrators, service meshes, ingress controllers and databases, without bespoke instrumentation.

OpenTelemetry’s Maturity: From Project to Pillar

OpenTelemetry has reached a pivotal point in its evolution. Initially viewed as an ambitious unification of OpenTracing and OpenCensus, it is now foundational infrastructure for a growing number of organisations. Several major themes emerged at KubeCon Europe 2025 that illustrate this transformation:

• Logs Signal Reaches General Availability: One of the most significant milestones was the general availability of log signal support. This means telemetry collection across metrics, traces and logs is now unified via the OTLP protocol and common schema. Structured logs can be directly correlated with spans and metrics, enabling a far more integrated approach to observability.
• Semantic Conventions 1.0: The introduction of Semantic Conventions 1.0 marks a step change in consistency across telemetry data. Attributes such as http.status_code, db.system, and k8s.pod.name are now applied uniformly across languages, runtimes and exporters, streamlining the development of dashboards, alerts, and automated analysis tools.
• Collector Performance Improvements: The OpenTelemetry Collector has received significant upgrades, with more efficient memory handling, better backpressure control, and modular pipeline capabilities. Performance benchmarks presented at the conference demonstrated throughput exceeding 1 million spans per minute per instance under load.
• WASI and eBPF: What’s Next: WebAssembly was a standout theme at KubeCon, highlighting its growing influence across the cloud native landscape, including in observability. Experimental use of WASI (WebAssembly System Interface) is enabling secure, portable data processing within the OpenTelemetry Collector, allowing teams to introduce dynamic logic such as filtering or enrichment. Combined with eBPF’s low-overhead, kernel-level insights, these technologies are helping to shape fast, flexible observability pipelines across Kubernetes environments.

Industry Alignment: Vendors and Platforms Converge

Perhaps the strongest indicator of OpenTelemetry’s maturity is the increasing alignment from major cloud providers and observability vendors:

• Istio and Linkerd now natively export telemetry via OTLP.
• Grafana Tempo and Loki have embraced OTel as a core ingestion mechanism.
• AWS Distro for OpenTelemetry (ADOT) supports full-fidelity tracing and metrics with native SDK support for Java, Python and Go, integrated into ECS and EKS by default.
• Azure Monitor’s OpenTelemetry Exporter is now GA, enabling direct OTLP ingestion and context-aware correlation.
• Google Cloud added OpenTelemetry tracing support to the Ops Agent, and GKE Autopilot includes OpenTelemetry Collector as a toggle in the UI.
• Red Hat OpenShift and VMware Tanzu now include curated OpenTelemetry configurations with pre-defined security policies for multi-tenant scenarios.

Why OpenTelemetry Matters

In a world increasingly dominated by microservices, containers and ephemeral infrastructure, observability is no longer optional, it’s essential. OpenTelemetry provides a vendor-neutral, open standard for capturing telemetry across any system, regardless of stack or deployment model.

Adopting OTel empowers teams to:

• Avoid vendor lock-in through open protocols (such as OTLP).
• Consolidate telemetry pipelines across services and signals.
• Scale observability efforts without spiralling operational complexity or cost.

Final Thoughts

KubeCon Europe 2025 confirmed what many had already suspected: OpenTelemetry has moved from a promising open source project to a fundamental pillar of cloud native operations. Its community is thriving, its adoption is accelerating, and its roadmap is focused and pragmatic.

For any organisation aiming to scale observability in an open, consistent, and future-ready way, OpenTelemetry is not just an option, it is the obvious choice.