> Markdown version of [/videos/100158-the-opentelemetry-mistakes-i-keep-seeing-and-how-to-stop-making-them?t=1435](https://www.wearedevelopers.com/videos/100158-the-opentelemetry-mistakes-i-keep-seeing-and-how-to-stop-making-them?t=1435). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md).

---

# The OpenTelemetry mistakes I keep seeing (and how to stop making them)

Is your OpenTelemetry setup leaking passwords or flooding your backend with noise? Learn how to permanently fix dangerous instrumentation anti-patterns today.

- **Speakers:** [Juraci Paixão Kröhling](https://www.wearedevelopers.com/@juraci-paixao-krohling)
- **Event:** World Congress 2026 Europe
- **Published:** July 9, 2026
- **Duration:** 27:02
- **URL:** https://www.wearedevelopers.com/videos/100158-the-opentelemetry-mistakes-i-keep-seeing-and-how-to-stop-making-them

## Summary

As AI-driven engineering and auto-generating agents take on coding tasks, comprehensive and accurate telemetry has become the primary source of truth for software in production. However, while frameworks like OpenTelemetry make starting easy, initial telemetry rollouts often calcify into persistent, noisy, and dangerous anti-patterns. A prevalent and severe consequence is inadvertently leaking sensitive data—ranging from personally identifiable information (PII) to JVM trust store passwords and signed URLs. Exposing this information to observability backends is "just like nuclear waste in the organic bin," posing immense security and compliance risks across environments.

Beyond data leaks, engineering teams frequently stumble over three distinct operational hurdles: over-instrumentation, under-instrumentation, and signal misalignment. Over-instrumentation manifests as traces inflated with thousands of microsecond spans, logging static assets, or placing heavy debug logs in the production hot path. Conversely, under-instrumentation leaves responders blind during early-morning alerts, highlighting the need to adopt a test-driven mindset for observability. Developers should map out failure scenarios beforehand and proactively embed the exact telemetry needed to answer critical debugging questions. Finally, teams often misuse signals—such as emitting thousands of textual log lines for a recurring error that should simply be a metric counter, or attempting to trace a distributed business transaction using fragmented logs instead of cohesive distributed traces.

Fixing these anti-patterns demands treating instrumentation with the same rigor as application security or testing. While auto-instrumentation tooling promises effortless observability, it frequently captures uncontrolled diagnostic scope and ships it externally. Developers must actively review their telemetry payloads inside tools like Grafana, Datadog, or Honeycomb to ensure they aren't shipping unusable junk or exposing vulnerabilities. Establishing a deliberate strategy where developers choose the most appropriate signal type—metrics for aggregate counters, traces for lifecycle flow, and actionable logs for pinpoint errors—results in sustainable observability that reliably accelerates root cause analysis.

**Keywords:** opentelemetry configuration, distributed tracing patterns, observability best practices, auto-instrumentation risks, PII data leaks, debugging production errors, JVM telemetry settings, telemetry signal selection, trace span cardinality, incident response readiness, AI generated code monitoring, infrastructure monitoring tools, application health checks, logs vs metrics, Grafana backends, Datadog integrations

## Chapters

1. **Why telemetry is critical for modern software development** (00:49) — Telemetry serves as the primary source of truth when developers delegate code writing to AI agents.
1. **Distinguishing between telemetry, monitoring, and observability** (04:56) — Telemetry provides raw data, monitoring answers known systems questions, and observability enables ad-hoc application investigations.
1. **Preventing sensitive data leaks in application telemetry** (08:27) — Capturing everything by default often leads to accidentally exposing passwords, email addresses, and personally identifiable information.
1. **Avoiding over-instrumentation and unnecessary tracing noise** (14:52) — Recording every microsecond function call or static asset request creates overwhelming system noise instead of actionable insights.
1. **Planning instrumentation around critical application failure scenarios** (17:37) — Anticipating late-night on-call emergencies guides the implementation of telemetry that effectively answers critical system questions.
1. **Choosing the correct telemetry signal for the job** (19:45) — Using logs for high-frequency application events obscures insights that simple metrics or distributed traces naturally provide.
1. **Key takeaways for building trustworthy application telemetry** (22:05) — Regularly reviewing raw production data reveals hidden data leaks and unnecessary noise before real incidents happen.
1. **Handling compliance, logging definitions, and AI agent output** (23:55) — Audience questions clarify GDPR compliance boundaries and identify dangerous default instrumentation patterns produced by AI code generators.

## Related Moments

- [Maintaining code quality with open standard observability and pipelines](https://www.wearedevelopers.com/videos/1706-the-ai-ready-stack-rethinking-the-engineering-org-of-the-future) (from "The AI-Ready Stack: Rethinking the Engineering Org of the Future")
- [Core concepts of application instrumentation and telemetry](https://www.wearedevelopers.com/videos/1232-observability-with-opentelemetry-elastic) (from "Observability with OpenTelemetry & Elastic")
- [Tracing agent telemetry with OpenTelemetry and Jaeger tools](https://www.wearedevelopers.com/videos/1601-one-ai-api-to-power-them-all) (from "One AI API to Power Them All")
- [Discovering incidents using logs, metrics, and traces](https://www.wearedevelopers.com/videos/680-handling-incidents-collaboratively-is-like-solving-a-rubix-cube) (from "Handling incidents collaboratively is like solving a rubix cube")
- [Final takeaways on application telemetry and tracing](https://www.wearedevelopers.com/videos/838-tips-techniques-and-common-pitfalls-debugging-kafka) (from "Tips, Techniques, and Common Pitfalls Debugging Kafka")
- [Exploring advanced observability stacks and distributed infrastructure challenges](https://www.wearedevelopers.com/videos/544-plan-ci-cd-on-the-enterprise-level) (from "Plan CI/CD on the Enterprise level!")

## Related Articles

- [MLops – Deploying, Maintaining And Evolving Machine Learning Models in Production](https://www.wearedevelopers.com/magazine/115-mlops-deploying-maintaining-and-evolving-machine-learning-models-in-production)
- [MLOps – What’s the deal behind it?](https://www.wearedevelopers.com/magazine/125-mlops-what-s-the-deal-behind-it)
- [Why Event-Driven Architecture Isn’t About Speed (and When You Actually Need It)](https://www.wearedevelopers.com/magazine/745-why-event-driven-architecture-isn-t-about-speed-and-when-you-actually-need-it)
- [Never delegate the understanding](https://www.wearedevelopers.com/magazine/749-never-delegate-the-understanding)

## Related Jobs

- [Senior Engineer, Infrastructure Platform](https://www.wearedevelopers.com/jobs/ext/328836-senior-engineer-infrastructure-platform) at **Intercom, Inc.**
- [Lead Software Engineer - Data Engineering](https://www.wearedevelopers.com/jobs/ext/2000968-lead-software-engineer-data-engineering) at **Dynatrace**
- [Engineer, Offensive Security Organization](https://www.wearedevelopers.com/jobs/ext/1992296-engineer-offensive-security-organization) at **Twilio**
- [Software Engineer, Platform Engineering (L2)](https://www.wearedevelopers.com/jobs/ext/1956829-software-engineer-platform-engineering-l2) at **Twilio**
- [Staff Software Engineer, Copilot Experiences](https://www.wearedevelopers.com/jobs/ext/164361-staff-software-engineer-copilot-experiences) at **GitHub**
- [Tribe Lead - ( Software) Engineering Centre of Excllence](https://www.wearedevelopers.com/jobs/ext/1475530-tribe-lead-software-engineering-centre-of-excllence) at **SD Worx**