Alerting When Customers Cannot Use the Service is a practical monitoring project for DevOps beginners with basic terminal familiarity. Northline Library looks open, but readers may still be unable to search or confirm a reservation. Build an alert around the customer outcome.
One Guided Lab connects a real search probe, a sustained-failure rule and actual local notifications. One independent Challenge transfers the method to reservation confirmation while excluding an unrelated failing service. Each unit has its own ready environment.
What You Will Learn
- Distinguish a reachable homepage or exporter from a successful customer journey.
- Select a content-aware HTTP probe and identify it with job and journey labels.
- Configure a short Prometheus rule that ignores brief failures and fires after sustained failure.
- Follow pending, firing and inactive states separately from notification delivery.
- Restore the customer path and confirm a real resolved webhook without stopping monitoring.
Who This Course Is For
This course suits new developers and DevOps learners who want a small, complete monitoring exercise.
Prerequisites: Basic terminal familiarity. HTTP, paths, YAML indentation, quoting, here-documents and rule syntax are explained before use. No Prometheus, application programming or email administration experience is required.
Learning environment: A browser-based Linux workstation, Library and Monitoring tabs are prepared with Prometheus, Blackbox Exporter, Alertmanager and a local webhook inbox. No local installation or external account is required. Books are synthetic and reservation previews are read-only. The website, interface labels and command output remain English; instructions and complete solutions support English plus eight localized languages.
Frequently Asked Questions
Do I need to build the application or install monitoring tools?
No. You edit two short configuration files and use a supplied controller to exercise real business faults. The controller changes the library's behavior; it does not set alert states.
Will the lab send email or messages to real people?
No. A local inbox receives real Alertmanager webhook deliveries inside the disposable training instance.
What makes the Challenge different?
You choose the reservation-specific path and labels, keeping a separate failing practice job outside your alert. Full solutions explain the reasoning and recovery.
Does this teach a complete production monitoring platform?
It focuses on one customer outcome, sustained failure and recovery. The short 20-second wait is for training. Production alert policies, high availability, dashboards, SLOs and external notification operations are outside scope.
