Monitoring & Alerting Setup for Kubernetes

This repository contains the setup for monitoring and alerting in a MicroK8s kubernetes cluster.

What is my goal here ( Monitoring & Alerting ) :

node-level metrics
pod-level metrics
pod logs
node-level logs ( dmesg, /var/log/messages, syslog )

Notification Method :

send alerts via email using Prometheus Alertmanager.

Components

Node Exporters: for collecting node-level metrics
Promtail-Loki: for collecting pod logs and node-level logs
Prometheus: for collecting and storing metrics
Alertmanager: for sending alerts via email

Storage :

This setup uses hostPath-based storage for observability, which can be changed according to your use case (If you decided not to go for hostpath based storage).

Loki :

This setup uses Loki (single binary/monolithic) for log collection and storage. Other options like Loki-scalable are available but not covered in this setup.

Kubernetes Distribution :

This setup uses MicroK8s as the Kubernetes distribution.

External URLs The following external URLs have been decided for:

Prometheus UI
Grafana
AlertManager
Promtail

Access All components are exposed using NodePort for externally public access. The format for accessing each component in my case is Public_ip_of_kmaster:NodePort.

vishalk17/observability-stack-kubernetes