Slide 3 of 6

Quick reference

The four levels at a glance

LevelKey questionPrimary toolDeep dive course
Level 1“Is my infrastructure healthy?”Integrations, Cloud Provider Olly, DB Olly, K8s MonitoringIntegrations, Data Sources, CP Olly, DB Olly, Fleet Management, Send Data
Level 2“Which service is the problem?”Application Observability (Service Inventory, RED)App Olly (coming soon)
Level 3“Why is this request slow?”Traces, Profiling, Frontend OllyTracing, Profiling (coming soon)
Level 4“What’s in my application logic?”OpenTelemetry SDKs, Custom MetricsCustom Instrumentation (coming soon)

Operational practices (cross-cutting)

PracticePrimary toolsWhen to focus
Alerting and IRMAlerting, SLOs, OnCall, IncidentOnce you have data at any level
Proactive testingSynthetic Monitoring, k6After Level 1 is stable
Platform managementAdaptive Telemetry, RBAC, SSOWhen scaling across teams
Grafana AssistantNatural language queries, troubleshootingAvailable at any level

Script

This slide is your cheat sheet. Bookmark it, screenshot it, print it. Whatever works for you.

Level 1: Is my infrastructure healthy? Use Integrations, Cloud Provider Observability, Database Observability, Kubernetes Monitoring.

Level 2: Which service is the problem? Use Application Observability with Service Inventory and RED metrics.

Level 3: Why is this request slow? Use Distributed Traces, Continuous Profiling, and Frontend Observability.

Level 4: What’s happening in my application logic? Use OpenTelemetry SDKs and custom metrics.

Operational practices (alerting, proactive testing, platform management) span all levels. Start using them as soon as you have data. And Grafana Assistant is available at any level to help you query data, troubleshoot issues, and build dashboards using natural language.

And remember the key insight: each level provides context for the next. While you can jump ahead if needed, building the foundation first helps you get the most value from advanced capabilities.