DEV108 · Monitoring, Logging and Incident Response
About Course
Know when your services are unhealthy before customers tell you, find the cause quickly and learn from every incident. You will build a local observability stack with metrics, dashboards, logs and alerts, then run a simulated incident from detection to review.
Who it is for
Developers, operations staff and IT leads responsible for keeping web applications available and fast.
Prerequisites
Complete DEV105 (DEV106 recommended) or be comfortable running multi-container applications with Docker Compose.
What you will be able to do
- Define service level indicators and objectives and use error budgets to guide decisions.
- Instrument an application and build Prometheus metrics, Grafana dashboards and actionable alerts.
- Collect structured logs and traces and use them to investigate a problem.
- Run an incident response process with clear roles, communication and a blameless post-incident review.
Lessons
- Lesson 1 · Observability fundamentals and service levels
- Lesson 2 · Metrics, dashboards and alerting
- Lesson 3 · Logs, traces and incident response
- Lesson 4 · Lab and assignment
Lab: Build an observability stack and handle an incident
You will start a Compose stack with a sample Node.js API, Prometheus, Grafana and Loki, expose request metrics from the API and build a dashboard of the golden signals. You will define an SLO and a burn-rate alert, then trigger a simulated failure, investigate it using metrics, logs and traces, and write the incident record.
Free tools: Docker Desktop with Docker Compose, Prometheus, Grafana, Loki with Grafana Alloy, Alertmanager, Uptime Kuma and the OpenTelemetry SDK for Node.js (all free and open source)
Assessed assignment
Design the monitoring and incident response setup for a small business web service, such as an online booking system for a beauty salon or a store's checkout API. Deliver two or three SLIs with SLOs and their error budgets, a Grafana dashboard covering latency, traffic, errors and saturation, at least three alert rules each linked to a short runbook, structured logging with a correlation id, and a one-page incident response plan with roles, severity levels and communication templates. Finish with a blameless post-incident review of the simulated incident from the lab, including a timeline and at least three action items.
| Criterion | Weight | What excellent work looks like |
|---|---|---|
| SLIs, SLOs and error budgets | 25% | Indicators reflect what users experience, objectives are realistic and the error budget is calculated correctly. |
| Dashboards and alerts | 30% | The dashboard shows the golden signals clearly, and alerts are symptom-based, actionable and linked to runbooks. |
| Logs and investigation | 20% | Logs are structured with a correlation id and the investigation uses metrics, logs and traces together. |
| Incident process and review | 25% | Roles, severities and communication are defined, and the review has a clear timeline, causes and owned actions. |
Portfolio piece
A working local observability stack with dashboards, alerts, runbooks and a documented post-incident review.
Next step
Continue with DEV109 · Final Project: An Automated Delivery Pipeline for a Web App · 8 hours · $56.00
Course Content
DEV108 · Monitoring, Logging and Incident Response (8 h)
-
Lesson 1 · Observability fundamentals and service levels
-
Lesson 2 · Metrics, dashboards and alerting
-
Lesson 3 · Logs, traces and incident response
-
Lesson 4 · Lab and assignment
-
DEV108 Assessment · Monitoring, Logging and Incident Response