Skip to content
hemzazPublic

About

Auto generate Grafana dashboards based on prometheus metrics endpoints

Resources

Stars

0 stars

Watchers

0 watching

Forks

 
 

Repository files navigation

lazydash

Auto generate Grafana dashboards based on prometheus metrics endpoints.
Allows for quick prototyping of dashboards.

Notes

Version: 1.0.0

  • Supports Counter, Gauge, Summary, and Histogram metrics
  • Advanced panel organization with label grouping and auto-correlation
  • Multiple visualization types (graphs, gauges, stats, tables, heatmaps)
  • Automatic alert generation
  • Folder organization in Grafana
  • Assumes metrics adhere to prometheus metrics naming conventions and standards e.g:
# HELP builder_builds_triggered_total Number of triggered image builds
# TYPE builder_builds_triggered_total counter
builder_builds_triggered_total 0"

Features

  • Generate dashboards directly from any Prometheus metrics endpoint
  • Post generated dashboards directly to Grafana via the API
  • Intelligent panel organization by label or correlation
  • Advanced visualization selection based on metric type
  • Auto panel legend format with all labels from prometheus
  • Panels use the metrics HELP field as the description
  • Override default query expressions and legend formats
  • Folder organization for dashboard management
  • Alert generation based on metric patterns

demo

---

Usage

usage: lazydash [<flags>]

Generate a Grafana dashboard from Prometheus metrics data via file, stdin or HTTP url

Flags:
  -h, --help             Show context-sensitive help (also try --help-long and --help-man).
      --version          Show application version.
  -f, --file=""          Parse metrics from file.
  -t, --title="Prometheus Dashboard"  
                         Dashboard title
      --description="Generated by Lazydash"  
                         Dashboard description
      --stdin            Read from stdin
      --url=""           Fetch Prometheus data from HTTP(S) url
  -p, --pretty           Print pretty indented JSON
  -g, --gauges           Render gauge values as gauge panel type instead of graph
      --table            Render legend as a table
      --set-counter-expr="sum(rate(:METRIC: [1m]))"  
                         Set custom meterics query expression for counter type metric
      --set-gauge-expr=":METRIC:"  
                         Set custom meterics query expression for gauge type metric
      --set-summary-expr=":METRIC:"  
                         Set custom meterics query expression for summary type metric
      --set-delimiter=":METRIC:"  
                         Set custom meterics delimiter used to insert metric name into expression
      --set-counter-legend="Job:[{{job}}]"  
                         Set the default counter panel legend format
      --set-gauge-legend="Job:[{{job}}]"  
                         Set the default counter panel legend format
  -H, --grafana-url=""   Set the grafana api url e.g http://grafana.example.com:3000
  -I, --insecure         Skip ssl certificate verification
  -T, --token=""         Set the grafana api token
      --folder=""        Set the Grafana folder name
      --folder-create    Create the folder if it doesn't exist
      --folder-description="Generated by Lazydash"  
                         Set the Grafana folder description
      --group-by= ...    Group panels by label
      --separate-rows    Create separate rows for each label group
      --panels-per-row=2 Number of panels per row (0 for auto)
      --auto-correlate   Automatically correlate related metrics
      --correlation-threshold=0.7  
                         Threshold for auto-correlation (0.0-1.0)
      --viz-counters=graph  
                         Visualization type for counters
      --viz-gauges=gauge  
                         Visualization type for gauges
      --viz-summaries=graph  
                         Visualization type for summaries
      --heatmap-for-histograms  
                         Use heatmap for histogram metrics
      --stat-for-gauges  Use stat panels for simple gauges
      --table-for-multilabels  
                         Use tables for metrics with many labels
      --generate-alerts  Automatically generate alerts for common metrics

Advanced Features

Organization

  • Label Grouping: Group panels by metric labels --group-by="job" --group-by="instance"
  • Folder Organization: Create and use Grafana folders --folder="My Dashboard" --folder-description="Generated dashboards"
  • Auto-correlation: Group related metrics together --auto-correlate --correlation-threshold=0.8
  • Vendor Grouping: Automatically detect and group vendor-specific metrics:
    • --vendor-detect - Enable vendor-specific metric detection
    • --group-by-vendor - Group metrics by vendor in the dashboard
    • --vendor-prefix="myvendor_" - Add custom vendor prefixes to detect
    • --juniper - Enable special handling for Juniper metrics
    • --cisco - Enable special handling for Cisco metrics

Visualization

  • Multiple Visualization Types: Choose from graphs, gauges, stats, tables, heatmaps
    • --viz-counters=graph
    • --viz-gauges=gauge
    • --viz-summaries=graph
  • Automatic Visualization Selection:
    • --heatmap-for-histograms - Use heatmaps for histogram metrics
    • --stat-for-gauges - Use stat panels for simple gauges
    • --table-for-multilabels - Use tables for metrics with many labels

Alerting

  • Auto-generated Alerts: --generate-alerts - Creates alert rules based on metric patterns

Examples

Pull metric types from prometheus HTTP endpoint and post to the grafana API

lazydash -p --url="http://localhost:9323/metrics" -H http://localhost:3000 -T "eyJrIjoiRzRTUGV1a2RWcjgzbklvVzdXenIySEhJWEJlSkx4UksiLCJuIjoidGVzdCIsImlkIjoxfQ==" --table

Advanced dashboard with grouping and visualization selection

lazydash -p --url=http://localhost:9323/metrics --group-by="job" --auto-correlate --heatmap-for-histograms --stat-for-gauges --generate-alerts > dashboard.json

Detect and group vendor-specific metrics (e.g., Juniper, Cisco)

lazydash -p --url=http://prometheus:9090/api/v1/query?query=up --vendor-detect --juniper --cisco --group-by-vendor --generate-alerts > network_devices.json

Process Juniper JTIMON metrics from a file

lazydash -f JuniperMetrics.txt --vendor-detect --juniper --group-by-vendor --pretty > juniper_dashboard.json

Override the counter expression and legend format

lazydash -p --url=http://localhost:9323/metrics --set-counter-expr="sum(rate(:METRIC: [1m])) by(instance)" --set-counter-legend="Instance:[{{instance}}]" > dash1.json

Pipe via curl

curl -s http://localhost:9323/metrics | lazydash -t "Demo" -p

Using file input

lazydash -f promdata.txt -p

Build and Install

Requirement: Go 1.21+

# Install from source
git clone https://github.com/hemzaz/lazydash.git
cd lazydash
go install ./cmd/lazydash

# Or use go install directly
go install github.com/hemzaz/lazydash/cmd/lazydash@latest

About

Auto generate Grafana dashboards based on prometheus metrics endpoints

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages