open-telemetry/opentelemetry-specification
Opentelemetry Specification
Este repositorio contiene las especificaciones para un estándar de datos de observabilidad estandarizado y un framework de instrumentación de telemetría. Define protocolos unificados para cómo se instrumentan, propagan y formatean los datos de telemetría para garantizar la interoperabilidad a través de sistemas distribuidos.
La especificación proporciona un enfoque neutral al proveedor para la observabilidad, permitiendo que los datos de telemetría se exporten a varios backends sin bloquearse en un solo proveedor o requerir cambios en el código fuente de la aplicación. Establece un protocolo consistente para capturar flujos de solicitudes a través de límites de servicio para correlacionar la actividad y analizar el rendimiento del sistema.
El framework cubre una amplia superficie de capacidades de monitoreo, incluyendo recolección de métricas del sistema, agregación de registros de aplicaciones y captura de trazas distribuidas. Define métodos tanto para la instrumentación automática como manual y especifica los mecanismos para la propagación de contexto y el enrutamiento de señales de telemetría a través de pipelines de datos.
Features
- Cloud Native Observability - Implements a vendor-neutral standard for collecting metrics, logs, and traces within cloud-native and distributed environments.
- Unified Observability Data Models - Defines a consistent data model that unifies metrics, logs, and traces as timestamped wide events.
- Distributed Tracing - Specifies the protocols for recording and visualizing the lifecycle of requests as they traverse distributed service architectures.
- Tracing Context Propagation - Implements mechanisms for injecting and extracting distributed tracing headers across network boundaries to maintain request correlation.
- Application Performance Monitoring - Provides a standardized approach for tracking runtime errors and performance bottlenecks in distributed software.
- Instrumentation APIs - Provides standardized interfaces for adding both manual and automatic observability markers to application code.
- Distributed Tracing - Defines methods for tracking requests across distributed systems using unique identifiers to analyze performance and execution flow.
- Vendor-Neutral Observability Interfaces - Provides standardized APIs that decouple application code from specific monitoring providers to prevent vendor lock-in.
- Zero-Code Instrumentation - Enables monitoring agents to be attached for telemetry collection without requiring modifications to application source code.
- Telemetry Data Pipelines - Provides a configurable layer for ingesting, parsing, and transforming raw observability data before export.
- OTLP Telemetry Routing - Routes observability data through processors and filters using OpenTelemetry standards.
- Distributed Log Aggregation - Collects and unifies timestamped event records from multiple distributed services for centralized monitoring.
- Telemetry Exporters - Pushes internal performance metrics and logs to various third-party observability platforms using a neutral standard.
- Instrumentation Libraries - Supplies libraries for adding custom tracking and tracing directly into application code via manual instrumentation.
- System Metrics Collection - Records system-level performance metrics over time to monitor resource usage and software health.
- Tracing And Profiling - Framework for cloud-native observability.
