Course Overview
Monitoring of cloud native applications and infrastructure.
Prometheus is a tool for metrics-based monitoring of applications and systems. The particular strength of Prometheus lies in its automatic service discovery, which ensures that all relevant targets are monitored, even in dynamic environments with frequent changes. Prometheus has a simple yet powerful data model and query language. Because of these features, Prometheus is commonly used for monitoring cloud and container environments as well as cloud native applications. This course offers a hands-on introduction to monitoring with Prometheus. Special attention is given to the integration of Prometheus and Kubernetes.
Who should attend
The training is aimed at developers and administrators as well as architects and decision-makers with a high level of technical interest.
Prerequisites
The participants have to master the basics of Linux system administration. They need to be able to proficiently use the command line, edit files on the command line, and manage system services. In addition, the participants need basic knowledge about the system properties and services to be monitored. Basic knowledge of Docker or Podman as well as Kubernetes is required for parts of the course.
Course Objectives
Participants learn how to monitor applications, services and infrastructure with Prometheus. This includes installing and configuring various exporters, service discovery and collecting metrics with Prometheus, as well as querying metrics via PromQL. In addition, the course covers creating dashboards with Grafana, defining alerting rules and configuring AlertManager, as well as an overview of Prometheus clustering, including Federation and Thanos.
Course Content
Introduction to monitoring with Prometheus
- Goals and methods of monitoring
- Architecture and components of Prometheus
- Selection of systems and properties to be monitored
Exporters
- Node exporter for monitoring individual systems
- Blackbox exporter for monitoring external systems
- Grok exporter and mtail for log file monitoring
- Push gateway for monitoring completed processes
- cAdvisor for monitoring containers
- kube-state-metrics for Kubernetes monitoring
Service discovery
- Static configuration
- File-based service discovery
- Service discovery via DNS
- Service discovery in IaaS clouds
- Service discovery in Kubernetes
- Relabeling, especially in IaaS clouds and Kubernetes
PromQL
- Metrics and labels
- Data types and their properties and functions
- Vector types and vector matching
- Recording rules and alerting rules
Alerting
- AlertManager
- Configuration of alerts
- Alert labels and alert routing
Dashboards and Grafana
- Prometheus Consoles
- Grafana
Operational aspects
- Meta monitoring
- Capacity planning
- Scaling and high availability
- Distributed Prometheus and Federation
- Hints for instrumenting custom applications
- Overview on Thanos