Datadog Telemetry Migration Made Easy

Bridging Datadog to OpenTelemetry

The article effectively highlights the cost pressures driving teams away from proprietary observability solutions like Datadog and presents a compelling technical solution via the OpenTelemetry Collector's Datadog receiver. The improvements to the receiver, such as accurate log record translation, resource attribute mapping, and trace/log correlation, are crucial for enabling a smooth transition. By allowing existing Datadog agents and SDKs to feed telemetry into the OpenTelemetry Collector pipeline, it significantly reduces the upfront migration burden, enabling parallel evaluation of new backends like ClickStack. This approach is particularly valuable for large organizations with substantial existing instrumentation, as it decouples the collection infrastructure migration from the backend platform migration, de-risking the overall process. The detailed worked example with ClickStack and the Hacker News demo further solidifies the practical applicability of the described solution.

However, it's important to acknowledge the stated limitations. The Datadog receiver is still in 'alpha'/'experimental' status, meaning it's not yet recommended for large-scale production migrations, especially concerning metrics, which require further testing. While the article focuses on the benefits for migration and evaluation, the long-term strategy for organizations adopting this approach should consider a gradual shift towards native OpenTelemetry instrumentation to fully leverage the ecosystem. The reliance on Datadog's native protocols for ingestion, even if translated, means that while the backend can change, the collection agents themselves are still configured to speak Datadog's language. This is a strategic intermediate step, but not the final destination for pure OpenTelemetry adoption. The article could also benefit from a more in-depth discussion on potential performance implications of the translation process within the Collector, especially under very high telemetry volumes, and perhaps a comparison of the translated telemetry fidelity against direct Datadog ingestion before the migration.

Key Points

• The OpenTelemetry Collector's Datadog receiver is significantly improved, enabling easier migration from Datadog.
• Existing Datadog agents and SDKs can send telemetry to the OpenTelemetry Collector, which translates it into the OpenTelemetry data model.
• This allows teams to evaluate new observability backends (like ClickStack) without immediately replacing all their instrumentation.
• Key improvements include better log record translation, resource attribute mapping, and trace/log correlation.
• The solution facilitates a phased migration by allowing parallel telemetry pipelines to Datadog and a new backend.
• While promising, the Datadog receiver is still experimental and not recommended for all large-scale production scenarios yet, particularly for metrics.

---

📖 Source: Migrate Datadog telemetry with the OpenTelemetry Collector