Observability should not be siloed into a single centralized team but treated as a shared responsibility across developers, SREs, QA, DevOps, and platform engineers. Developers should own code-based OpenTelemetry instrumentation since they know their applications best, while platform/SRE teams manage OTel Collector infrastructure. CI/CD pipelines also benefit from observability, with an emerging OTel CI/CD SIG providing a shared standard. Organizations should have an 'observability practices team' focused on standards, tooling governance, and cross-team advocacy rather than hands-on instrumentation or infrastructure management. Avoiding tool sprawl and operating without silos is key to effective observability at scale.
Table of contents
Observability is everyone’s responsibilityEnsuring adherence to observability practicesFinal thoughtsSort: