KubeCon + CloudNativeCon + Open Source Summit + AI_dev China 2024

The cloud's vast landscape is characterized by a diversity of applications that employ different tracing protocols, each tailored to specific telemetry collection tools and requirements. This diversity results in an ecosystem where maintaining consistent and reliable traces across disparate systems is a significant challenge.
To address this, our contribution enhances OpenTelemetry with an adaptive approach to propagate multi-protocol trace signals. Building on this implementation, we have developed an adaptable and extendable trace propagation framework, facilitating a more seamless trace propagation process and ensuring instant compatibility with a diverse range of cloud services.
During this talk, we'll dive deep into the design and implementation of this feature and introduce how this mechanism achieves out-of-the-box functionality within our APM services.

云计算的广阔领域以各种应用为特征，这些应用采用不同的跟踪协议，每种协议都针对特定的遥测收集工具和需求进行了定制。这种多样性导致了一个生态系统，在这个生态系统中，跨不同系统保持一致且可靠的追踪是一个重大挑战。

为了解决这一问题，我们的贡献在OpenTelemetry基础上增强了自适应的跟踪信号传播方法。在这一实现的基础上，我们开发了一个灵活且可扩展的跟踪传播框架，促进了更无缝的追踪传播过程，并确保与各种云服务的即时兼容性。

在这次演讲中，我们将深入探讨这一特性的设计和实现，并介绍这一机制如何在我们的应用性能管理（APM）服务中实现即插即用的功能。