OpenTelemetry with PHP is easier than you think!
summary
Repository with Docker Compose and ready demos exists and auto‑instrumentation uses a PHP 8 extension hook to attach spans and metrics. Local stacks shown include Zipkin, Jaeger and Apprace and deployment options compare agent per host versus centralized collectors. A real surprise: an early run generates 20 million spans a minute while vendor limits are 6M so metrics plus sampling (errors plus ~1% others, aim <5%) become the practical, cost‑saving priority. Trace_test and SQL commenter are used for assertions and correlating slow queries.
AI-written summary · about
chapters (17)
- 0:00Introduction
- 1:07What is OpenTelemetry
- 3:55OpenTelemetry vs APM tools
- 6:28Demonstration of tools
- 8:43Signals and context
- 14:03Collector deployment options
- 16:53Instrumenting PHP code
- 18:19PHP extension and hooks
- 22:17Metrics importance
- 26:30Production deployment lessons
- 30:10Manual vs automatic tracing
- 34:27Hooks for metrics
- 35:30Advanced tools like Trace Test
- 37:40Future of OpenTelemetry
- 38:23Key lessons and best practices
- 40:11PHP configuration tips
- 41:39Final recommendations
related talks
46mMonitoring a PHP application with OpenTelemetry
35mOpenTelemetry for Busy People
38mOpenTelemetry: toward a standard for monitoring your applications
17mOpenTelemetry Looks Good To Me: Loki, Grafana, Tempo, Mimir
22mOpenTelemetry – An Observability Framework for Cloud-Native Software
02h 02mHow To Troubleshoot and Monitor Applications using OpenTelemetry
34mOpenTelemetry for Dummies: Instrumenting Go Apps
38mThe State of OpenTelemetry