Learning from the visible past to accelerate the observable future
Observability is a highly visible topic in DevOps and Cloud Native discussions. But does observability need a bigger vision or broader applicability to break through to the next level? Can past precedents point a way to accommodate current and future technical trends? Metrics and logs are well understood, and distributed tracing has surfaced on the global developer consciousness. OpenTelemetry is a very good starting point for the next iteration, but it only addresses a subset of the problem space. Several important, but under-represented, observability challenges are identified which a reference model needs to address. Lessons learned from building and using an observability platform are discussed. Based on this, an observability reference model is proposed to outline the whole problem space and provide the North Star for the journey ahead.