Speaker
Description
In a real-world production environment at large scale, where performance is crucial, it is very useful for software developers to understand how their code behaves and to precisely identify where bottlenecks may potentially occur.
To obtain such information, developers can instrument their software, embedding probes and meters to collect data remotely. This concept, known in the industry as telemetry, permits to trace interesting code execution paths and precisely measure their time
We have built and deployed a stack of several services to ingest different types of telemetry data, collected by instrumenting some software products we develop, enabling the real-time display of key product indicators (KPIs) on custom dashboards built in Grafana.
In this contribution we present the described collector stack and how we have instrumented StoRM Tape, StorRM WebDAV and INDIGO IAM using the popular framework OpenTelemetry.
