Monitor Config Sync with Prometheus

The page describes how to send metrics from Config Sync to Prometheus.

This page explains how to use Prometheus to view Config Sync metrics. For other ways to export metrics, see Monitor Config Sync with Cloud Monitoring or Monitor Config Sync with custom monitoring.

Config Sync automatically collects and exports metrics to Prometheus. You can configure Cloud Monitoring to pull custom metrics from Prometheus. Then you can see custom metrics in both Prometheus and Monitoring. For more information, see Using Prometheus in the GKE documentation.

Scrape the metrics

All Prometheus metrics are available for scraping at port 8675. Before you can scrape metrics, you need to configure your cluster for Prometheus in one of two ways. Either:

  • Follow the Prometheus documentation to configure your cluster for scraping, or

  • Use the Prometheus Operator along with the following manifests, which scrape all Config Sync metrics every 10 seconds.

    1. Create a temporary directory to hold the manifest files.

      mkdir config-sync-monitor
cd config-sync-monitor

    2. Download the Prometheus Operator manifest from the CoreOS repository. repository, using the curl command:

      curl -o bundle.yaml https://raw.githubusercontent.com/coreos/prometheus-operator/master/bundle.yaml

      This manifest is configured to use the default namespace, which is not recommended. The next step modifies the configuration to use a namespace called monitoring instead. To use a different namespace, substitute it where you see monitoring in the remaining steps.

    3. Create a file to update the namespace of the ClusterRoleBinding in the bundle above.

      # patch-crb.yaml
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
  name: prometheus-operator
subjects:
- kind: ServiceAccount
  name: prometheus-operator
  namespace: monitoring # we are patching from default namespace

    4. Create a kustomization.yaml file that applies the patch and modifies the namespace for other resources in the manifest.

      # kustomization.yaml
resources:
- bundle.yaml

namespace: monitoring

patchesStrategicMerge:
- patch-crb.yaml

    5. Create the monitoring namespace if one does not exist. You can use a different name for the namespace, but if you do, also change the value of namespace in the YAML manifests from the previous steps.

      kubectl create namespace monitoring

    6. Apply the Kustomize manifest using the following commands:

      kubectl apply -k .

until kubectl get customresourcedefinitions servicemonitors.monitoring.coreos.com ; \
do date; sleep 1; echo ""; done

      The second command blocks until the CRDs are available on the cluster.

    7. Create the manifest for the resources necessary to configure a Prometheus server which scrapes metrics from Config Sync.

      # config-sync-monitoring.yaml
apiVersion: v1
kind: ServiceAccount
metadata:
  name: prometheus-config-sync
  namespace: monitoring
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRole
metadata:
  name: prometheus-config-sync
rules:
- apiGroups: [""]
  resources:
  - nodes
  - services
  - endpoints
  - pods
  verbs: ["get", "list", "watch"]
- apiGroups: [""]
  resources:
  - configmaps
  verbs: ["get"]
- nonResourceURLs: ["/metrics"]
  verbs: ["get"]
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
  name: prometheus-config-sync
roleRef:
  apiGroup: rbac.authorization.k8s.io
  kind: ClusterRole
  name: prometheus-config-sync
subjects:
- kind: ServiceAccount
  name: prometheus-config-sync
  namespace: monitoring
---
apiVersion: monitoring.coreos.com/v1
kind: Prometheus
metadata:
  name: config-sync
  namespace: monitoring
  labels:
    prometheus: config-sync
spec:
  replicas: 2
  serviceAccountName: prometheus-config-sync
  serviceMonitorSelector:
    matchLabels:
      prometheus: config-management
  alerting:
    alertmanagers:
    - namespace: default
      name: alertmanager
      port: web
  resources:
    requests:
      memory: 400Mi
---
apiVersion: v1
kind: Service
metadata:
  name: prometheus-config-sync
  namespace: monitoring
  labels:
    prometheus: config-sync
spec:
  type: NodePort
  ports:
  - name: web
    nodePort: 31900
    port: 9190
    protocol: TCP
    targetPort: web
  selector:
    prometheus: config-sync
--- 
---

    8. Apply the manifest using the following commands:

      kubectl apply -f config-sync.yaml

until kubectl rollout status statefulset/prometheus-config-sync -n monitoring; \
do sleep 1; done

      The second command blocks until the Pods are running.

    9. You can verify the installation by forwarding the web port of the Prometheus server to your local machine.

      kubectl -n monitoring port-forward svc/prometheus-config-sync 9190

      You can now access the Prometheus web UI at http://localhost:9190.

    10. Remove the temporary directory.

      cd ..
rm -rf config-sync-monitor

Available Prometheus metrics

Config Sync collects the following metrics and makes them available to Prometheus. The Labels column lists all labels that are applicable to each metric. Metrics without labels represent a single measurement over time while metrics with labels represent multiple measurements, one for each combination of label values.

If this table becomes out of sync, you can filter metrics by prefix in the Prometheus user interface. All of the metrics start with the prefix config_sync_.

Name Type Labels Description
config_sync_api_duration_seconds_bucket Histogram status, operation Latency distribution of API server calls (distributed into buckets by duration of each cycle)
config_sync_api_duration_seconds_count Histogram status, operation Latency distribution of API server calls (ignoring duration)
config_sync_api_duration_seconds_sum Histogram status, operation Sum of the durations of all API server calls
config_sync_apply_duration_seconds_bucket Histogram commit, status Latency distribution of applying resources declared from source of truth to a cluster (distributed into buckets by duration of each cycle)
config_sync_apply_duration_seconds_count Histogram commit, status Latency distribution of applying resources declared from source of truth to a cluster (ignoring duration)
config_sync_apply_duration_seconds_sum Histogram commit, status Sum of the durations of all the latency of applying resources declared from source of truth to a cluster
config_sync_apply_operations_total Counter operation, status, controller Number of operations that have been performed to sync resources from source of truth to a cluster
config_sync_cluster_scoped_resource_count Gauge resourcegroup Number of cluster scoped resources in a ResourceGroup
config_sync_crd_count Gauge resourcegroup Number of CRDs in a ResourceGroup
config_sync_declared_resources Gauge commit Number of declared resources parsed from Git
config_sync_internal_errors_total Counter source Number of internal errors triggered by Config Sync. Metric might not appear if no internal error has happened
config_sync_kcc_resource_count Gauge resourcegroup Number of Config Connector resources in a ResourceGroup
config_sync_last_apply_timestamp Gauge commit, status Timestamp of the most recent apply operation
config_sync_last_sync_timestamp Gauge commit, status Timestamp of the most recent sync from Git
config_sync_parser_duration_seconds_bucket Histogram status, trigger, source Latency distribution of different stages involved in syncing from source of truth to a cluster
config_sync_parser_duration_seconds_count Histogram status, trigger, source Latency distribution of different stages involved in syncing from source of truth to a cluster (ignoring duration)
config_sync_parser_duration_seconds_sum Histogram status, trigger, source Sum of the latencies of different stages involved in syncing from source of truth to a cluster
config_sync_pipeline_error_observed Gauge name, reconciler, component Status of RootSync and RepoSync custom resources. A value of 1 indicates a failure
config_sync_ready_resource_count Gauge resourcegroup Total number of ready resources in a ResourceGroup
config_sync_reconcile_duration_seconds_bucket Histogram status Latency distribution of reconcile events handled by the reconciler manager (distributed into buckets by duration of each call)
config_sync_reconcile_duration_seconds_count Histogram status Latency distribution of reconcile events handled by the reconciler manager (ignoring duration)
config_sync_reconcile_duration_seconds_sum Histogram status Sum of the durations of all latency of reconcile events handled by the reconciler manager
config_sync_reconciler_errors Gauge component, errorclass Number of errors encountered while syncing resources from the source of the truth to a cluster
config_sync_remediate_duration_seconds_bucket Histogram status Latency distribution of remediator reconciliation events (distributed into buckets by duration)
config_sync_remediate_duration_seconds_count Histogram status Latency distribution of remediator reconciliation events (ignoring duration)
config_sync_remediate_duration_seconds_sum Histogram status Sum of the durations of all latency of remediator reconciliation events
config_sync_resource_count Gauge resourcegroup Number of resources tracked by a ResourceGroup
config_sync_resource_conflicts_total Counter commit Number of resource conflicts resulting from a mismatch between the cached resources and cluster resources. Metric might not appear if no resource conflict has happened
config_sync_resource_fights_total Counter Number of resources that are being synced too frequently. Metric might not appear if no resource fight has happened
config_sync_resource_group_total Gauge Number of ResourceGroup CRs
config_sync_resource_ns_count Gauge resourcegroup Number of namespaces used by resources in a ResourceGroup
config_sync_rg_reconcile_duration_seconds_bucket. Histogram stallreason Time distribution of reconciling a ResourceGroup CR (distributed into buckets by duration)
config_sync_rg_reconcile_duration_seconds_count Histogram stallreason Time distribution of reconciling a ResourceGroup CR (ignoring duration)
config_sync_rg_reconcile_duration_seconds_sum Histogram stallreason Sum of the all time reconciling a ResourceGroup CR
config_sync_kustomize_build_latency_bucket Histogram Latency distribution of kustomize build execution time (distributed into buckets by duration of each operation)
config_sync_kustomize_build_latency_count Histogram Latency distribution of kustomize build execution time (ignoring duration)
config_sync_kustomize_build_latency_sum Histogram Sum of all kustomize build execution time
config_sync_kustomize_ordered_top_tier_metrics Gauge top_tier_field Usage of Resources, Generators, SecretGenerator, ConfigMapGenerator, Transformers, and Validators
config_sync_kustomize_builtin_transformers Gauge k8s_builtin_transformer Usage of built-in transformers related to kubernetes object metadata
config_sync_kustomize_resource_count Gauge Number of resources outputted by kustomize build
config_sync_kustomize_field_count Gauge field_name Number of times a particular field is used in the kustomization files
config_sync_kustomize_patch_count Gauge patch_field Number of patches in the fields patches, patchesStrategicMerge, and patchesJson6902
config_sync_kustomize_base_count Gauge base_source Number of remote and local bases
kustomize_deprecating_field_count Gauge deprecating_field Usage of fields that may become deprecated
kustomize_simplification_adoption_count Gauge simplification_field Usage of simplification transformers images, replicas, and replacements
kustomize_helm_inflator_count Gauge helm_inflator Usage of helm in kustomize, whether by the built-in fields or the custom function

Example debugging procedures for Prometheus

The following examples illustrate some patterns for using Prometheus metrics, object status fields, and object annotations to detect and diagnose problems related to Config Sync. These examples show how you can start with high level monitoring that detects a problem and then progressively refine your search to drill down and diagnose the root cause of the problem.

Query configs by status

The reconciler process provides high-level metrics that give useful insight into an overall view of how Config Sync is operating on the cluster. You can see if any errors have occurred, and can even set up alerts for them.

config_sync_reconciler_errors

Query metrics by reconciler

If you are using Config Sync RootSync and RepoSync APIs, then you can monitor the RootSync and RepoSync objects. The RootSync and RepoSync objects are instrumented with high-level metrics that give you useful insight into how Config Sync is operating on the cluster. Almost all metrics are tagged by the reconciler name, so you can see if any errors have occurred and can set up alerts for them in Prometheus.

See the full list of available metric labels for filtering.

In Prometheus, you can use the following filters for the RootSyncs or RepoSyncs:

# Querying RootSync
config_sync_reconciler_errors{configsync_sync_name=ROOT_SYNC_NAME}

# Querying RepoSync
config_sync_reconciler_errors{configsync_sync_name=REPO_SYNC_NAME}

Query import and sync operations by status

In Prometheus, you could use the following queries:

# Check for errors that occurred when sourcing configs.
config_sync_reconciler_errors{component="source"}

# Check for errors that occurred when syncing configs to the cluster.
config_sync_reconciler_errors{component="sync"}

You can also check the metrics for the source and sync processes themselves:

config_sync_parser_duration_seconds{status="error"}
config_sync_apply_duration_seconds{status="error"}
config_sync_remediate_duration_seconds{status="error"}

Monitor resources with Google Cloud Managed Service for Prometheus

Google Cloud Managed Service for Prometheus is Google Cloud's fully managed multi-cloud solution for Prometheus metrics. It supports two modes for data collection: managed collection (the recommended mode) or self-deployed data collection. Complete the following steps to set up monitoring Config Sync with Google Cloud Managed Service for Prometheus in the managed collection mode.

  1. Enable Managed Prometheus on your cluster by following the instructions on Set up managed collection.

  2. Save the following sample manifest as pod-monitoring-config-sync-monitoring.yaml. This manifest configures a PodMonitoring resource to scrape the Config Sync metrics on port 8675 of the otel-collector-* Pod under the config-management-monitoring namespace. The PodMonitoring resource uses a Kubernetes label selector to find the otel-collector-* Pod.

    apiVersion: monitoring.googleapis.com/v1
kind: PodMonitoring
metadata:
  name: config-sync-monitoring
  namespace: config-management-monitoring
spec:
  selector:
    matchLabels:
      app: opentelemetry
      component: otel-collector
  endpoints:
  - port: 8675
    interval: 10s

  3. Apply the manifest to the cluster: 

    kubectl apply -f pod-monitoring-config-sync-monitoring.yaml

  4. Verify that your Prometheus data is being exported using the Cloud Monitoring Metrics Explorer page in the Google Cloud console following the instructions on Managed Service for Prometheus data in Cloud Monitoring.