Datadog APM & Distributed Tracing Fundamentals study plan
Datadog APM & Distributed Tracing Fundamentals · Give each week a clear focus.
8 weeks · About 6 hours per week
APM fundamentals
Define the core APM vocabulary (trace, span, service, resource, operation name) and map each term to where it appears in the Datadog APM UI
Describe how distributed traces are assembled from spans and explain parent-child relationships and context propagation across service boundaries
Explain the difference between a trace, a span, and a service map entry, and identify which one you would inspect for a given symptom
Summarize how trace ingestion, sampling, and retention affect what you can later query, without relying on any external materials
Your checkpoint
A one-page glossary plus a hand-drawn trace diagram showing a three-service request with parent-child spans and context propagation annotated
APM fundamentals & Application instrumentation with Datadog
Compare head-based and retention-based sampling and state the trade-offs for visibility versus cost
Identify the instrumentation entry points for a language you use (agent, tracer library, environment variables) and describe what each configures
List the environment variables and agent settings that control service naming, environment tagging, and version tagging
Explain how unified service tagging (service, env, version) connects traces to other telemetry
Your checkpoint
A configuration checklist for one language runtime covering tracer install, agent connection, and unified service tagging, validated against a running local demo app
Application instrumentation with Datadog
Instrument a sample multi-service application so that traces appear end to end with correct service names
Add manual spans around a business-critical code path and set meaningful resource and operation names
Add custom tags and span tags to capture business context (for example customer tier or order value)
Verify context propagation between at least two services by inspecting a single trace in the UI
Your checkpoint
A working instrumented demo repository with a screenshot-free written trace walkthrough showing one trace spanning two or more services with custom tags
Application instrumentation with Datadog & Insight discovery with Datadog APM
Configure trace retention filters and custom sampling rules for a chosen service and justify the choices
Use the trace search and analytics interface to filter spans by tag, resource, and duration
Distinguish between live search and indexed (retained) trace querying and explain when each applies
Interpret the service list, service page, and service map views to locate a slow or erroring service
Your checkpoint
A written investigation report that isolates a deliberately introduced latency regression using only trace search and the service map, with the query used and findings recorded
Insight discovery with Datadog APM
Create a custom span-based metric from an existing span tag and explain its retention behavior
Build a monitor or alert on a trace-derived signal such as error rate or p95 latency for a specific service
Use the trace view to compare a slow trace against a baseline trace and identify the dominant span
Interpret error tracking signals in APM, including error rates, error messages, and stack traces
Your checkpoint
A saved custom metric plus an APM monitor definition, each documented with the query, threshold rationale, and expected alert behavior
Visualizing insights
Build a dashboard that combines service latency, error rate, and throughput for a chosen service
Add a trace analytics widget or service map widget that links a visual panel back to underlying traces
Configure a dashboard template variable so the dashboard can be scoped by environment or service
Explain how to move from a dashboard anomaly to the specific trace that caused it
Your checkpoint
An exported dashboard definition (JSON or documented panel list) with at least four panels, one template variable, and a written drill-down path from panel to trace
Troubleshooting applications using APM & Insight discovery with Datadog APM
Diagnose a missing-trace scenario by checking agent status, tracer logs, and sampling configuration in order
Diagnose a broken-trace scenario where context propagation fails between two services
Use the deployment and version tracking views to correlate a regression with a version change
Explain how to read a flame graph or span waterfall to locate the slowest span in a request
Your checkpoint
A troubleshooting runbook covering five failure modes (no traces, partial traces, wrong service name, high latency, elevated errors) with the diagnostic steps and expected evidence for each
APM fundamentals & Application instrumentation with Datadog & Insight discovery with Datadog APM & Visualizing insights & Troubleshooting applications using APM
Rehearse the full workflow from symptom to root cause using trace search, service map, and dashboards
Review weak areas identified in earlier deliverables and close gaps with targeted practice
Explain sampling, unified service tagging, and retention filters from memory without notes
Time-box a set of practice scenarios to build pacing for the exam format
Your checkpoint
A timed practice set of at least 40 self-written scenario questions across all five domains, scored, with a one-page list of remaining weak areas and a remediation note for each
Milestones
Week 2: Can explain head-based versus retention-based sampling and list the unified service tagging variables without notes
Week 4: Has a working instrumented multi-service app and has successfully isolated an injected latency regression using only trace search and the service map
Week 6: Has a functioning dashboard with a template variable and can articulate the drill-down path from panel to trace
Week 8: Scores at least 80 percent on the timed practice set and has a written remediation note for every missed domain