loki-stack

Loki Stack Helm Chart

Observability stack - Loki for logs, Grafana for visualization, Prometheus for metrics, and Grafana Alloy for log collection (alternative to deprecated Promtail).

DOCS: https://rishang.github.io/helm-chart/loki-stack/introduction

Overview

This Helm chart deploys a complete observability stack on Kubernetes, providing:

• Loki: Log aggregation system for storing and querying logs
• Grafana: Visualization and dashboards for logs and metrics
• Prometheus: Metrics collection and monitoring via kube-prometheus-stack
• Grafana Alloy: Modern log collection agent (replaces deprecated Promtail)

Prerequisites

• Kubernetes 1.22+
• Helm 3.0+
• Persistent storage (optional, but recommended for production)

Installation

Add the repository (if applicable)

helm repo add common-charts https://rishang.github.io/helm-chart/charts
helm repo update

Install the chart

helm install loki-stack common-charts/loki-stack --namespace monitoring --create-namespace

Install with custom values

helm install loki-stack . --namespace monitoring --create-namespace --values custom-values.yaml

Configuration

Key Configuration Options

Parameter	Description	Default
loki.enabled	Enable Loki log aggregation	true
loki.deploymentMode	Loki deployment mode	SingleBinary
loki.singleBinary.persistence.enabled	Enable persistent storage for Loki	false
alloy.enabled	Enable Grafana Alloy log collector	true
kube-prometheus-stack.enabled	Enable Prometheus and Grafana	true
kube-prometheus-stack.grafana.enabled	Enable Grafana dashboard	true

Example: Enable Persistence

loki:
  singleBinary:
    persistence:
      enabled: true
      size: 10Gi
      storageClass: standard

Example: Disable Components

# Disable Prometheus stack, keep only Loki and Alloy
kube-prometheus-stack:
  enabled: false

Accessing Services

Grafana Dashboard

Port-forward to access Grafana:

kubectl port-forward -n monitoring svc/loki-stack-grafana 3000:80

Default credentials:

• Username: admin
• Password: Run kubectl get secret -n monitoring loki-stack-grafana -o jsonpath="{.data.admin-password}" | base64 --decode

Loki Query API

Port-forward to access Loki directly:

kubectl port-forward -n monitoring svc/loki-stack-loki 3100:3100

Query logs: http://localhost:3100/loki/api/v1/query

Architecture

This chart deploys:

1. Loki (SingleBinary mode): Stores and indexes logs
2. Grafana Alloy (DaemonSet): Collects logs from /var/log/pods on each node
3. Grafana: Provides UI for log exploration and visualization
4. Prometheus: Collects metrics from Kubernetes and applications
5. Alertmanager: Handles alerts from Prometheus

Log Collection

Grafana Alloy is configured to:

• Mount /var/log from each node
• Discover and tail logs from /var/log/pods/*/*/*.log
• Parse Kubernetes metadata (namespace, pod, container)
• Forward logs to Loki with proper labels

Upgrading

helm upgrade loki-stack . --namespace monitoring --values custom-values.yaml

Uninstallation

helm uninstall loki-stack --namespace monitoring

Note: PersistentVolumeClaims are not automatically deleted. Remove them manually if needed:

kubectl delete pvc -n monitoring -l app.kubernetes.io/instance=loki-stack

Testing

Generate manifest for validation:

helm template loki-stack . --namespace monitoring --values test/values.yaml > loki-stack-manifest.yaml

Dependencies

This chart depends on:

• Loki
• Grafana Alloy
• kube-prometheus-stack

Support

For issues and questions:

• Chart Issues: GitHub Issues
• Loki Documentation: https://grafana.com/docs/loki/
• Alloy Documentation: https://grafana.com/docs/alloy/