Foray / Observability EngineerObservability Engineer roadmap See inside running systems: metrics, logs and traces, dashboards and alerts that earn trust, SLOs that turn data into decisions, and the platform and budget that keep it all running.
13 topics · 54 nodes · 46 covered · 4 partly · 4 planned · updated 2026-10-17
Sign in with your Mycelium account to see which nodes you have finished.
Show Whole map Gaps only 1. Core
Systems foundations You can only observe what you understand. The operating-system and network basics behind most signals.
4 / 4 covered
Linux signals: CPU, memory, disk Covered A troubleshooting method Covered Network and HTTP basics Covered Containers and Kubernetes basics optional Covered 2. Core
Observability concepts The vocabulary and the mindset before the tools.
4 / 4 covered
What observability is Covered What to measure Covered Metrics, logs and traces Covered Observability habits Covered 3. Core
Metrics with Prometheus The de facto standard for time-series metrics in cloud-native systems.
5 / 6 covered
How Prometheus works Covered PromQL Covered Cardinality and cost Covered Instrumentation and exporters Covered Service discovery and relabeling Covered Alternatives to Prometheus pick one Partly covered 4. Core
Dashboards with Grafana Turn data into something a tired human can read.
2 / 3 covered
Grafana dashboards Covered Dashboards that answer questions Partly covered Observability as code Covered 5. Core
Alerting and on-call Alerts should wake people for the right reasons.
5 / 5 covered
Alerting basics Covered Alerting on SLOs Covered Runbooks Covered Designing alerts Covered On-call practice Covered 6. Core
Logging The record of what happened, in detail.
3 / 4 covered
Logging basics Covered System logs Covered Centralised logging with Loki Covered Elasticsearch, OpenSearch and vendors pick one Planned 7. Core
Tracing and OpenTelemetry Follow a request across services.
3 / 4 covered
Distributed tracing Covered Telemetry pipelines Covered OpenTelemetry Covered Tempo, Jaeger and vendors pick one Planned 8. Recommended
Kubernetes and cloud monitoring Where most telemetry comes from today.
5 / 5 covered
Monitoring Kubernetes Covered Dependencies and the outside view Covered Web server and proxy metrics optional Covered Cloud monitoring Covered Service mesh observability optional Covered 9. Core
SLOs and reliability Observability exists to answer one question: is it good enough for users?
4 / 4 covered
SLIs, SLOs and SLAs Covered SLOs in practice Covered Error budgets Covered Capacity planning Covered 10. Core
Incident response and investigation When something breaks, observability is the investigator's toolkit.
4 / 4 covered
Investigating incidents Covered Blameless postmortems Covered Incident management Covered Testing what you can see optional Covered 11. Recommended
Scaling and running the platform The observability system is production too.
4 / 4 covered
Scaling Prometheus Covered Controlling telemetry cost Covered Running the observability platform Covered Security and privacy in telemetry Covered 12. Optional
Advanced signals Beyond the three pillars.
2 / 4 covered
Profiling and eBPF Covered Commercial APM platforms pick one Planned Business and frontend observability Covered AIOps and anomaly detection optional Planned 13. Recommended
Automating observability Make instrumentation and monitoring repeatable.
1 / 3 covered
Python for ops tooling Covered Deploying agents and exporters optional Partly covered Monitoring through infrastructure as code Partly covered © 2026 Konstantinos Maglaras
