The onboarding process
The need for distributed tracing and visibility of all parts of the system comes from the complexity of modern applications. For example, we need to know how a serverless environment interacts with cloud storage to debug configuration issues or optimize performance. Or, maybe we want to know why certain requests fail in the downstream service without asking someone to help.
To make the most of distributed tracing, we have to onboard the whole system (or at least a significant part of it), making sure all services create correlated and coherent telemetry, write it to a place where different teams can access it, and reuse the same tooling to do analysis.
So, implementing an observability solution is an organization-wide effort, and it makes sense to start with a pilot project instrumenting a small part of the system. Let’s outline its scope and goals.
The pilot phase
The goal of this project is to get hands-on experience with observability, discover...
