Monitoring Enterprise

Weave GitOps Enterprise provides monitoring telemetry and tooling for metrics and profiling. WGE generates Prometheus metrics for monitoring both performance and business operations.

Setup

The following configuration options are available for you to configure monitoring:

---
apiVersion: helm.toolkit.fluxcd.io/v2beta1
kind: HelmRelease
metadata:
  name: weave-gitops-enterprise
  namespace: flux-system
spec:
  values:
    monitoring:
      enabled: true # enable it if you want to expose a monitoring server
      service:
        name: monitoring
        port: 8080 # port to expose the monitoring server
      metrics:
        enabled: true # enable it to expose a prometheus metrics endpoint in `/metrics`
      profiling:
        enabled: false # enable it to expose a pprof debug endpoint  `/debug/pprof`

caution

The monitoring server holds private services, so you probably won't need to expose anything beyond your cluster. If you must, ensure that it is properly secured.

Get Started with Monitoring

This setup follows the same monitoring approach as Flux and is based on Prometheus Operator. Adapt it to your context as needed.

1. Enable the monitoring server with the metrics endpoint.
2. Install Kube Prometheus Stack.

Expand to see manifest contents

apiVersion: source.toolkit.fluxcd.io/v1
kind: GitRepository
metadata:
  name: weave-gitops-quickstart
  namespace: flux-system
spec:
  interval: 10m0s
  ref:
    branch: main
  url: https://github.com/weaveworks/weave-gitops-quickstart
---
apiVersion: v1
kind: Namespace
metadata:
  name: monitoring
---
apiVersion: kustomize.toolkit.fluxcd.io/v1
kind: Kustomization
metadata:
  name: kube-prometheus-stack
  namespace: flux-system
spec:
  interval: 10m0s
  sourceRef:
    kind: GitRepository
    name: weave-gitops-quickstart
  path: ./monitoring/kube-prometheus-stack
  prune: true
  targetNamespace: monitoring
  wait: true

3. Deploy Weave GitOps Monitoring Config

Expand to see manifest contents

apiVersion: kustomize.toolkit.fluxcd.io/v1
kind: Kustomization
metadata:
  name: monitoring-config
  namespace: flux-system
spec:
  interval: 10m0s
  sourceRef:
    kind: GitRepository
    name: weave-gitops-quickstart
  path: ./monitoring/weave-gitops
  dependsOn:
    - name: kube-prometheus-stack
  prune: true
  targetNamespace: monitoring

4. See the dashboards in Grafana. You can filter by tags flux or weave-gitops.

Dashboards

Weave GitOps Overview

Monitor Weave GitOps golden signals for API server and controllers:

Weave GitOps Runtime

Monitor Weave GitOps Go runtime metrics like memory usage, memory heap, and Goroutines, among others.

Explorer

You can also monitor Explorer golden signals.

Profiling

During operations, profiling is useful for gaining a deeper understanding of how Weave GitOps runtime behaves. Given that Weave GitOps is written in Go, profiling happens through pprof. It is exposed as a web endpoint by pprof http.

Get Started with Profiling

1. Enable the monitoring server with the profiling endpoint.
2. Navigate to your monitoring server URL to the /debug/pprof path where the pprof web interface is exposed.

Go here for more info on using pprof.