/

Console

• English
• Deutsch
• Español
• Español – América Latina
• Français
• Indonesia
• Italiano
• Português
• Português – Brasil
• עברית
• 中文 – 简体
• 中文 – 繁體
• 日本語
• 한국어

Sign in

• Google Cloud Observability

Start free

Get started with managed collection

gcloud components list

kubectl create ns NAMESPACE_NAME

Managed Prometheus Bulk Cluster Enablement dashboard

You can do the following by using the Managed Prometheus Bulk Cluster Enablement dashboard in Cloud Monitoring.

• Determine whether Managed Service for Prometheus is enabled on your clusters and whether you are using managed or self-deployed collection.
• Enable managed collection on clusters in your project.
• View other information about your clusters.

To view the Managed Prometheus Bulk Cluster Enablement dashboard, do the following:

1. In the Google Cloud console, go to the dashboard Dashboards page:

   Go to Dashboards

   If you use the search bar to find this page, then select the result whose subheading is Monitoring.

2. Use the filter bar to search for the Managed Prometheus Bulk Cluster Enablement entry, then select it.

To enable managed collection on one or more GKE clusters by using the Managed Prometheus Bulk Cluster Enablement dashboard, do the following:

1. Select the checkbox for each GKE cluster on which you want to enable managed collection.

2. Select Enable Selected.

gcloud config set project PROJECT_ID

gcloud container clusters create CLUSTER_NAME --zone ZONE --enable-managed-prometheus

Disable managed collection

If you want to disable managed collection on your clusters, then you can use one of the following methods:

gcloud config set project PROJECT_ID

gcloud container clusters create CLUSTER_NAME --zone ZONE --no-enable-managed-prometheus

Enable managed collection: non-GKE Kubernetes

If you are running in a non-GKE environment, then you can enable managed collection using the following:

• The kubectl CLI.

• VMware or bare metal on-premises deployments running version 1.12 or newer.

kubectl apply -f https://raw.githubusercontent.com/GoogleCloudPlatform/prometheus-engine/v0.17.2/manifests/setup.yaml

kubectl apply -f https://raw.githubusercontent.com/GoogleCloudPlatform/prometheus-engine/v0.17.2/manifests/operator.yaml

Deploy the example application

The example application emits the example_requests_total counter metric and the example_random_numbers histogram metric (among others) on its metrics port. The manifest for the application defines three replicas.

To deploy the example application, run the following command:

kubectl -n NAMESPACE_NAME apply -f https://raw.githubusercontent.com/GoogleCloudPlatform/prometheus-engine/v0.17.2/examples/example-app.yaml

Configure a PodMonitoring resource

To ingest the metric data emitted by the example application, Managed Service for Prometheus uses target scraping. Target scraping and metrics ingestion are configured using Kubernetes custom resources. The managed service uses PodMonitoring custom resources (CRs).

A PodMonitoring CR scrapes targets only in the namespace the CR is deployed in. To scrape targets in multiple namespaces, deploy the same PodMonitoring CR in each namespace. You can verify the PodMonitoring resource is installed in the intended namespace by running kubectl get podmonitoring -A.

For reference documentation about all the Managed Service for Prometheus CRs, see the prometheus-engine/doc/api reference.

The following manifest defines a PodMonitoring resource, prom-example, in the NAMESPACE_NAME namespace. The resource uses a Kubernetes label selector to find all pods in the namespace that have the label app.kubernetes.io/name with the value prom-example. The matching pods are scraped on a port named metrics, every 30 seconds, on the /metrics HTTP path.

apiVersion: monitoring.googleapis.com/v1
kind: PodMonitoring
metadata:
  name: prom-example
spec:
  selector:
    matchLabels:
      app.kubernetes.io/name: prom-example
  endpoints:
  - port: metrics
    interval: 30s

To apply this resource, run the following command:

kubectl -n NAMESPACE_NAME apply -f https://raw.githubusercontent.com/GoogleCloudPlatform/prometheus-engine/v0.17.2/examples/pod-monitoring.yaml

Your managed collector is now scraping the matching pods. You can view the status of your scrape target by enabling the target status feature.

To configure horizontal collection that applies to a range of pods across all namespaces, use the ClusterPodMonitoring resource. The ClusterPodMonitoring resource provides the same interface as the PodMonitoring resource but does not limit discovered pods to a given namespace.

If you are running on GKE, then you can do the following:

• To query the metrics ingested by the example application using PromQL in Cloud Monitoring, see Query using Cloud Monitoring.
• To query the metrics ingested by the example application using Grafana, see Query using Grafana or any Prometheus API consumer.
• To learn about filtering exported metrics and adapting your prom-operator resources, see Additional topics for managed collection.

If you are running outside of GKE, then you need to create a service account and authorize it to write your metric data, as described in the following section.

Provide credentials explicitly

When running on GKE, the collecting Prometheus server automatically retrieves credentials from the environment based on the node's service account. In non-GKE Kubernetes clusters, credentials must be explicitly provided through the OperatorConfig resource in the gmp-public namespace.

gcloud config set project PROJECT_ID

gcloud iam service-accounts create gmp-test-sa

gcloud projects add-iam-policy-binding PROJECT_ID\
  --member=serviceAccount:gmp-test-sa@PROJECT_ID.iam.gserviceaccount.com \
  --role=roles/monitoring.metricWriter

gcloud iam service-accounts keys create gmp-test-sa-key.json \
  --iam-account=gmp-test-sa@PROJECT_ID.iam.gserviceaccount.com

kubectl -n gmp-public create secret generic gmp-test-sa \
  --from-file=key.json=gmp-test-sa-key.json

Open the OperatorConfig resource for editing:

kubectl -n gmp-public edit operatorconfig config

apiVersion: monitoring.googleapis.com/v1
kind: OperatorConfig
metadata:
  namespace: gmp-public
  name: config
collection:
  credentials:
    name: gmp-test-sa
    key: key.json

Save the file and close the editor. After the change is applied, the pods are re-created and start authenticating to the metric backend with the given service account.

Additional topics for managed collection

This section describes how to do the following:

• Enable the target status feature for easier debugging.
• Configure target scraping using Terraform.
• Filter the data you export to the managed service.
• Scrape Kubelet and cAdvisor metrics.
• Convert your existing prom-operator resources for use with the managed service.
• Run managed collection outside of GKE.

Enabling the target status feature

Managed Service for Prometheus provides a way to check whether your targets are being properly discovered and scraped by the collectors. This target status report is meant to be a tool for debugging acute problems. We strongly recommend only enabling this feature to investigate immediate issues. Leaving target status reporting on in large clusters might cause the operator to run out of memory and crash loop.

You can check the status of your targets in your PodMonitoring or ClusterPodMonitoring resources by setting the features.targetStatus.enabled value within the OperatorConfig resource to true, as shown in the following:

apiVersion: monitoring.googleapis.com/v1
kind: OperatorConfig
metadata:
  namespace: gmp-public
  name: config
features:
  targetStatus:
    enabled: true

After a few seconds, the Status.Endpoint Statuses field appears on every valid PodMonitoring or ClusterPodMonitoring resource, when configured.

If you have a PodMonitoring resource with the name prom-example in the NAMESPACE_NAME namespace, then you can check the status by running the following command:

kubectl -n NAMESPACE_NAME describe podmonitorings/prom-example

The output looks like the following:

API Version:  monitoring.googleapis.com/v1
Kind:         PodMonitoring
...
Status:
  Conditions:
    ...
    Status:                True
    Type:                  ConfigurationCreateSuccess
  Endpoint Statuses:
    Active Targets:       3
    Collectors Fraction:  1
    Last Update Time:     2023-08-02T12:24:26Z
    Name:                 PodMonitoring/custom/prom-example/metrics
    Sample Groups:
      Count:  3
      Sample Targets:
        Health:  up
        Labels:
          Cluster:                     CLUSTER_NAME
          Container:                   prom-example
          Instance:                    prom-example-589ddf7f7f-hcnpt:metrics
          Job:                         prom-example
          Location:                    REGION
          Namespace:                   NAMESPACE_NAME
          Pod:                         prom-example-589ddf7f7f-hcnpt
          project_id:                  PROJECT_ID
        Last Scrape Duration Seconds:  0.020206416
        Health:                        up
        Labels:
          ...
        Last Scrape Duration Seconds:  0.054189485
        Health:                        up
        Labels:
          ...
        Last Scrape Duration Seconds:  0.006224887

The output includes the following status fields:

• Status.Conditions.Status is true when Managed Service for Prometheus acknowledges and processes the PodMonitoring or ClusterPodMonitoring.
• Status.Endpoint Statuses.Active Targets shows the number of scrape targets that Managed Service for Prometheus counts on all collectors for this PodMonitoring resource. In the example application, the prom-example deployment has three replicas with a single metric target, so the value is 3. If there are unhealthy targets, the Status.Endpoint Statuses.Unhealthy Targets field appears.
• Status.Endpoint Statuses.Collectors Fraction shows a value of 1 (meaning 100%) if all of the managed collectors are reachable by Managed Service for Prometheus.
• Status.Endpoint Statuses.Last Update Time shows the last updated time. When the last update time is significantly longer than your desired scrape interval time, the difference might indicate issues with your target or cluster.
• Status.Endpoint Statuses.Sample Groups field shows sample targets grouped by common target labels injected by the collector. This value is useful for debugging situations where your targets are not discovered. If all targets are healthy and being collected, then the expected value for the Health field is up and the value for the Last Scrape Duration Seconds field is the usual duration for a typical target.

For more information about these fields, see the Managed Service for Prometheus API document.

Any of the following might indicate a problem wth your configuration:

• There is no Status.Endpoint Statuses field in your PodMonitoring resource.
• The value of the Last Scrape Duration Seconds field is too old.
• You see too few targets.
• The value of the Health field indicates that the target is down.

For more information about debugging target discovery issues, see Ingestion-side problems in the troubleshooting documentation.

Configuring an authorized scrape endpoint

If your scrape target requires authorization, you can set up the collector to use the correct authorization type and provide any relevant secrets.

Google Cloud Managed Service for Prometheus supports the following authorization types:

• mTLS
• BasicAuth
• HTTP Authorization Header
• OAuth 2

mTLS

mTLS is commonly configured within zero trust environments, such as Istio service mesh or Cloud Service Mesh.

To enable scraping endpoints secured using mTLS, set the Spec.Endpoints[].Scheme field in your PodMonitoring resource to https. While not recommended, you can set the Spec.Endpoints[].tls.insecureSkipVerify field in your PodMonitoring resource to true to skip verifying the certificate authority. Alternatively, you can configure Managed Service for Prometheus to load certificates and keys from secret resources.

For example, the following Secret resource contains keys for the client (cert), private key (key), and certificate authority (ca) certificates:

kind: Secret
metadata:
  name: secret-example
stringData:
  cert: ********
  key: ********
  ca: ********

Grant the Managed Service for Prometheus collector permission to access that Secret resource:

apiVersion: rbac.authorization.k8s.io/v1
kind: Role
metadata:
  name: secret-example-read
rules:
- resources:
  - secrets
  apiGroups: [""]
  verbs: ["get", "list", "watch"]
  resourceNames: ["secret-example"]
---
apiVersion: rbac.authorization.k8s.io/v1
kind: RoleBinding
metadata:
  name: gmp-system:collector:secret-example-read
  namespace: default
roleRef:
  name: secret-example-read
  kind: Role
  apiGroup: rbac.authorization.k8s.io
subjects:
- name: collector
  namespace: gmp-system
  kind: ServiceAccount

On GKE Autopilot clusters, this looks like:

apiVersion: rbac.authorization.k8s.io/v1
kind: Role
metadata:
  name: secret-example-read
rules:
- resources:
  - secrets
  apiGroups: [""]
  verbs: ["get", "list", "watch"]
  resourceNames: ["secret-example"]
---
apiVersion: rbac.authorization.k8s.io/v1
kind: RoleBinding
metadata:
  name: gmp-system:collector:secret-example-read
  namespace: default
roleRef:
  name: secret-example-read
  kind: Role
  apiGroup: rbac.authorization.k8s.io
subjects:
- name: collector
  namespace: gke-gmp-system
  kind: ServiceAccount

To configure a PodMonitoring resource that uses the prior Secret resource, modify your resource to add a scheme and tls section:

apiVersion: monitoring.googleapis.com/v1
kind: PodMonitoring
metadata:
  name: prom-example
spec:
  selector:
    matchLabels:
      app.kubernetes.io/name: prom-example
  endpoints:
  - port: metrics
    interval: 30s
    scheme: https
    tls:
      ca:
        secret:
          name: secret-example
          key: ca
      cert:
        secret:
          name: secret-example
          key: cert
      key:
        secret:
          name: secret-example
          key: key

For reference documentation about all the Managed Service for Prometheus mTLS options, see the API reference documentation.

BasicAuth

To enable scraping endpoints secured using BasicAuth, set the Spec.Endpoints[].BasicAuth field in your PodMonitoring resource with your username and password. For other HTTP Authorization Header types, see HTTP Authorization Header.

For example, the following Secret resource contains a key to store the password:

kind: Secret
metadata:
  name: secret-example
stringData:
  password: ********

Grant the Managed Service for Prometheus collector permission to access that Secret resource:

apiVersion: rbac.authorization.k8s.io/v1
kind: Role
metadata:
  name: secret-example-read
rules:
- resources:
  - secrets
  apiGroups: [""]
  verbs: ["get", "list", "watch"]
  resourceNames: ["secret-example"]
---
apiVersion: rbac.authorization.k8s.io/v1
kind: RoleBinding
metadata:
  name: gmp-system:collector:secret-example-read
  namespace: default
roleRef:
  name: secret-example-read
  kind: Role
  apiGroup: rbac.authorization.k8s.io
subjects:
- name: collector
  namespace: gmp-system
  kind: ServiceAccount

On GKE Autopilot clusters, this looks like:

apiVersion: rbac.authorization.k8s.io/v1
kind: Role
metadata:
  name: secret-example-read
rules:
- resources:
  - secrets
  apiGroups: [""]
  verbs: ["get", "list", "watch"]
  resourceNames: ["secret-example"]
---
apiVersion: rbac.authorization.k8s.io/v1
kind: RoleBinding
metadata:
  name: gmp-system:collector:secret-example-read
  namespace: default
roleRef:
  name: secret-example-read
  kind: Role
  apiGroup: rbac.authorization.k8s.io
subjects:
- name: collector
  namespace: gke-gmp-system
  kind: ServiceAccount

To configure a PodMonitoring resource that uses the prior Secret resource and a username of foo, modify your resource to add a basicAuth section:

apiVersion: monitoring.googleapis.com/v1
kind: PodMonitoring
metadata:
  name: prom-example
spec:
  selector:
    matchLabels:
      app.kubernetes.io/name: prom-example
  endpoints:
  - port: metrics
    interval: 30s
    basicAuth:
      username: foo
      password:
        secret:
          name: secret-example
          key: password

For reference documentation about all the Managed Service for Prometheus BasicAuth options, see the API reference documentation.

HTTP Authorization Header

To enable scraping endpoints secured using HTTP Authorization Headers, set the Spec.Endpoints[].Authorization field in your PodMonitoring resource with the type and credentials. For BasicAuth endpoints, use the BasicAuth configuration instead.

For example, the following Secret resource contains a key to store the credentials:

kind: Secret
metadata:
  name: secret-example
stringData:
  credentials: ********

Grant the Managed Service for Prometheus collector permission to access that Secret resource:

apiVersion: rbac.authorization.k8s.io/v1
kind: Role
metadata:
  name: secret-example-read
rules:
- resources:
  - secrets
  apiGroups: [""]
  verbs: ["get", "list", "watch"]
  resourceNames: ["secret-example"]
---
apiVersion: rbac.authorization.k8s.io/v1
kind: RoleBinding
metadata:
  name: gmp-system:collector:secret-example-read
  namespace: default
roleRef:
  name: secret-example-read
  kind: Role
  apiGroup: rbac.authorization.k8s.io
subjects:
- name: collector
  namespace: gmp-system
  kind: ServiceAccount

On GKE Autopilot clusters, this looks like:

apiVersion: rbac.authorization.k8s.io/v1
kind: Role
metadata:
  name: secret-example-read
rules:
- resources:
  - secrets
  apiGroups: [""]
  verbs: ["get", "list", "watch"]
  resourceNames: ["secret-example"]
---
apiVersion: rbac.authorization.k8s.io/v1
kind: RoleBinding
metadata:
  name: gmp-system:collector:secret-example-read
  namespace: default
roleRef:
  name: secret-example-read
  kind: Role
  apiGroup: rbac.authorization.k8s.io
subjects:
- name: collector
  namespace: gke-gmp-system
  kind: ServiceAccount

To configure a PodMonitoring resource that uses the prior Secret resource and a type of Bearer, modify your resource to add an authorization section:

apiVersion: monitoring.googleapis.com/v1
kind: PodMonitoring
metadata:
  name: prom-example
spec:
  selector:
    matchLabels:
      app.kubernetes.io/name: prom-example
  endpoints:
  - port: metrics
    interval: 30s
    authorization:
      type: Bearer
      credentials:
        secret:
          name: secret-example
          key: credentials

For reference documentation about all the Managed Service for Prometheus HTTP Authorization Header options, see the API reference documentation.

OAuth 2

To enable scraping endpoints secured using OAuth 2, you must set the Spec.Endpoints[].OAuth2 field in your PodMonitoring resource.

For example, the following Secret resource contains a key to store the client secret:

kind: Secret
metadata:
  name: secret-example
stringData:
  clientSecret: ********

Grant the Managed Service for Prometheus collector permission to access that Secret resource:

apiVersion: rbac.authorization.k8s.io/v1
kind: Role
metadata:
  name: secret-example-read
rules:
- resources:
  - secrets
  apiGroups: [""]
  verbs: ["get", "list", "watch"]
  resourceNames: ["secret-example"]
---
apiVersion: rbac.authorization.k8s.io/v1
kind: RoleBinding
metadata:
  name: gmp-system:collector:secret-example-read
  namespace: default
roleRef:
  name: secret-example-read
  kind: Role
  apiGroup: rbac.authorization.k8s.io
subjects:
- name: collector
  namespace: gmp-system
  kind: ServiceAccount

On GKE Autopilot clusters, this looks like:

apiVersion: rbac.authorization.k8s.io/v1
kind: Role
metadata:
  name: secret-example-read
rules:
- resources:
  - secrets
  apiGroups: [""]
  verbs: ["get", "list", "watch"]
  resourceNames: ["secret-example"]
---
apiVersion: rbac.authorization.k8s.io/v1
kind: RoleBinding
metadata:
  name: gmp-system:collector:secret-example-read
  namespace: default
roleRef:
  name: secret-example-read
  kind: Role
  apiGroup: rbac.authorization.k8s.io
subjects:
- name: collector
  namespace: gke-gmp-system
  kind: ServiceAccount

To configure a PodMonitoring resource that uses the prior Secret resource with a client ID of foo and token URL of example.com/token, modify your resource to add an oauth2 section:

apiVersion: monitoring.googleapis.com/v1
kind: PodMonitoring
metadata:
  name: prom-example
spec:
  selector:
    matchLabels:
      app.kubernetes.io/name: prom-example
  endpoints:
  - port: metrics
    interval: 30s
    oauth2:
      clientID: foo
      clientSecret:
        secret:
          name: secret-example
          key: password
      tokenURL: example.com/token

For reference documentation about all the Managed Service for Prometheus OAuth 2 options, see the API reference documentation.

Configuring target scraping using Terraform

You can automate the creation and management of PodMonitoring and ClusterPodMonitoring resources by using the kubernetes_manifest Terraform resource type or the kubectl_manifest Terraform resource type, either of which lets you specify arbitrary custom resources.

For general information about using Google Cloud with Terraform, see Terraform with Google Cloud.

Filter exported metrics

If you collect a lot of data, you might want to prevent some time series from being sent to Managed Service for Prometheus to keep down costs. You can do this by using Prometheus relabeling rules with a keep action for an allowlist or a drop action for a denylist. For managed collection, this rule goes in the metricRelabeling section of your PodMonitoring or ClusterPodMonitoring resource.

For example, the following metric relabeling rule will filter out any metric that begins with foo_bar_, foo_baz_, or foo_qux_:

  metricRelabeling:
  - action: drop
    regex: foo_(bar|baz|qux)_.+
    sourceLabels: [__name__]

The Cloud Monitoring Metrics Management page provides information that can help you control the amount you spend on billable metrics without affecting observability. The Metrics Management page reports the following information:

• Ingestion volumes for both byte- and sample-based billing, across metric domains and for individual metrics.
• Data about labels and cardinality of metrics.
• Number of reads for each metric.
• Use of metrics in alerting policies and custom dashboards.
• Rate of metric-write errors.

You can also use the Metrics Management page to exclude unneeded metrics, eliminating the cost of ingesting them. For more information about the Metrics Management page, see View and manage metric usage.

For additional suggestions on how to lower your costs, see Cost controls and attribution.

Scraping Kubelet and cAdvisor metrics

The Kubelet exposes metrics about itself as well as cAdvisor metrics about containers running on its node. You can configure managed collection to scrape Kubelet and cAdvisor metrics by editing the OperatorConfig resource. For instructions, see the exporter documentation for Kubelet and cAdvisor.

Convert existing prometheus-operator resources

You can usually convert your existing prometheus-operator resources to Managed Service for Prometheus managed collection PodMonitoring and ClusterPodMonitoring resources.

For example, the ServiceMonitor resource defines monitoring for a set of services. The PodMonitoring resource serves a subset of the fields served by the ServiceMonitor resource. You can convert a ServiceMonitor CR to a PodMonitoring CR by mapping the fields as described in the following table:

The following is a sample ServiceMonitor CR; the content in bold type is replaced in the conversion, and the content in italic type maps directly:

apiVersion: monitoring.coreos.com/v1
kind: ServiceMonitor
metadata:
  name: example-app
spec:
  selector:
    matchLabels:
      app: example-app
  endpoints:
  - targetPort: web
    path: /stats
    interval: 30s
  targetLabels:
  - foo

The following is the analogous PodMonitoring CR, assuming that your service and its pods are labeled with app=example-app. If this assumption does not apply, then you need to use the label selectors of the underlying Service resource.

The content in bold type has been replaced in the conversion:

apiVersion: monitoring.googleapis.com/v1
kind: PodMonitoring
metadata:
  name: example-app
spec:
  selector:
    matchLabels:
      app: example-app
  endpoints:
  - port: web
    path: /stats
    interval: 30s
  targetLabels:
    fromPod:
    - from: foo # pod label from example-app Service pods.
      to: foo

You can always continue to use your existing prometheus-operator resources and deployment configs by using self-deployed collectors instead of managed collectors. You can query metrics sent from both collector types, so you might want to use self-deployed collectors for your existing Prometheus deployments while using managed collectors for new Prometheus deployments.

Reserved labels

Managed Service for Prometheus automatically adds the following labels to all metrics collected. These labels are used to uniquely identify a resource in Monarch:

• project_id: The identifier of the Google Cloud project associated with your metric.
• location: The physical location (Google Cloud region) where the data is stored. This value is typically the region of your GKE cluster. If data is collected from an AWS or on-premises deployment, then the value might be the closest Google Cloud region.
• cluster: The name of the Kubernetes cluster associated with your metric.
• namespace: The name of the Kubernetes namespace associated with your metric.
• job: The job label of the Prometheus target, if known; might be empty for rule-evaluation results.
• instance: The instance label of the Prometheus target, if known; might be empty for rule-evaluation results.

While not recommended when running on Google Kubernetes Engine, you can override the project_id, location, and cluster labels by adding them as args to the Deployment resource within operator.yaml. If you use any reserved labels as metric labels, Managed Service for Prometheus automatically relabels them by adding the prefix exported_. This behavior matches how upstream Prometheus handles conflicts with reserved labels.

Compress configurations

If you have many PodMonitoring resources, you might run out of ConfigMap space. To fix this, enable gzip compression in your OperatorConfig resource:

  apiVersion: monitoring.googleapis.com/v1
  kind: OperatorConfig
  metadata:
    namespace: gmp-public
    name: config
  features:
    config:
      compression: gzip

Enable vertical pod autoscaling (VPA) for managed collection

If you are encountering Out of Memory (OOM) errors for the collector pods in your cluster or if the default resource requests and limits for the collectors otherwise don't meet your needs, then you can use vertical pod autoscaling to dynamically allocate resources.

When you set the field scaling.vpa.enabled: true on the OperatorConfig resource, the operator deploys a VerticalPodAutoscaler manifest in the cluster that allows the resource requests and limits of the collector pods to be set automatically, based on usage.

To enable VPA for collector pods in Managed Service for Prometheus, run the following command:

kubectl -n gmp-public patch operatorconfig/config -p '{"scaling":{"vpa":{"enabled":true}}}' --type=merge

If the command completes successfully, then the operator sets up vertical pod autoscaling for the collector pods. Out Of Memory errors result in an immediate increase to the resource limits. If there are no OOM errors, then the first adjustment to the resource requests and limits of the collector pods typically occurs within 24 hours.

You might receive this error when attempting to enable VPA:

vertical pod autoscaling is not available - install vpa support and restart the operator

To resolve this error, you need to first enable vertical pod autoscaling at the cluster level:

1. Go to the Kubernetes Engine - Clusters page in the Google Cloud console.

   In the Google Cloud console, go to the Kubernetes clusters page:

   Go to Kubernetes clusters

   If you use the search bar to find this page, then select the result whose subheading is Kubernetes Engine.

2. Select the cluster you want to modify.

3. In the Automation section, edit the value of the Vertical Pod Autoscaling option.

4. Select the Enable Vertical Pod Autoscaling checkbox, and then click Save changes. This change restarts your cluster. The operator restarts as a part of this process.

5. Retry the following command: kubectl -n gmp-public patch operatorconfig/config -p '{"scaling":{"vpa":{"enabled":true}}}' --type=merge to enable VPA for Managed Service for Prometheus.

To confirm that the OperatorConfig resource is edited successfully, open it using the command kubectl -n gmp-public edit operatorconfig config. If successful, your OperatorConfig includes the following section in bold:

apiVersion: monitoring.googleapis.com/v1
kind: OperatorConfig
metadata:
  namespace: gmp-public
  name: config
scaling:
  vpa:
    enabled: true

If you have already enabled vertical pod autoscaling at the cluster level and are still seeing the vertical pod autoscaling is not available - install vpa support and restart the operator error, then the gmp-operator pod might need to re-evaluate the cluster configuration. Do one of the following:

• If you are running a Standard cluster, run the following command to recreate the pod:

  kubectl -n gmp-system rollout restart deployment/gmp-operator
  

  After the gmp-operator pod has restarted, follow the steps above to patch the OperatorConfig once again.

• If you are running an Autopilot cluster, then you can't restart the gmp-operator pod manually. When you enable VPA for the managed collectors in an Autopilot cluster, VPA automatically evicts and recreates the collector pods to apply the new resource requests; no cluster restart is required. If you see the vertical pod autoscaling is not available error after enabling VPA, or encounter other issues with VPA activation for Managed Service for Prometheus, then contact support.

Vertical pod autoscaling works best when ingesting steady numbers of samples, divided equally across nodes. If the metrics load is irregular or spiky, or if metrics load varies greatly between nodes, VPA might not be an efficient solution.

For more information, see vertical pod autoscaling in GKE.

Configure statsd_exporter and other exporters that report metrics centrally

If you use the statsd_exporter for Prometheus, Envoy for Istio, the SNMP exporter, the Prometheus Pushgateway, kube-state-metrics, or you otherwise have a similar exporter that intermediates and reports metrics on behalf of other resources running in your environment, then you need to make some small changes for your exporter to work with Managed Service for Prometheus.

For instructions on configuring these exporters, see this note in the Troubleshooting section.

Teardown

To disable managed collection deployed using gcloud or the GKE UI, you can do either of the following:

• Run the following command:

  gcloud container clusters update CLUSTER_NAME --disable-managed-prometheus
  

• Use the GKE UI:

  1. Select Kubernetes Engine in the Google Cloud console, then select Clusters.

  2. Locate the cluster for which you want to disable managed collection and click its name.

  3. On the Details tab, scroll down to Features and change the state to Disabled by using the edit button.

To disable managed collection deployed by using Terraform, specify enabled = false in the managed_prometheus section of the google_container_cluster resource.

To disable managed collection deployed by using kubectl, run the following command:

kubectl delete -f https://raw.githubusercontent.com/GoogleCloudPlatform/prometheus-engine/v0.17.2/manifests/operator.yaml

Disabling managed collection causes your cluster to stop sending new data to Managed Service for Prometheus. Taking this action does not delete any existing metrics data already stored in the system.

Disabling managed collection also deletes the gmp-public namespace and any resources within it, including any exporters installed in that namespace.

Run managed collection outside of GKE

In GKE environments, you can run managed collection without further configuration. In other Kubernetes environments, you need to explicitly provide credentials, a project-id value to contain your metrics, a location value (Google Cloud region) where your metrics will be stored, and a cluster value to save the name of the cluster in which the collector is running.

As gcloud does not work outside of Google Cloud environments, you need to deploy using kubectl instead. Unlike with gcloud, deploying managed collection using kubectl does not automatically upgrade your cluster when a new version is available. Remember to watch the releases page for new versions and manually upgrade by re-running the kubectl commands with the new version.

You can provide a service account key by modifying the OperatorConfig resource within operator.yaml as described in Provide credentials explicitly. You can provide project-id, location, and cluster values by adding them as args to the Deployment resource within operator.yaml.

We recommend choosing project-id based on your planned tenancy model for reads. Pick a project to store metrics in based on how you plan to organize reads later with metrics scopes. If you don't care, you can put everything into one project.

For location, we recommend choosing the nearest Google Cloud region to your deployment. The further the chosen Google Cloud region is from your deployment, the more write latency you'll have and the more you'll be affected by potential networking issues. You might want to consult this list of regions across multiple clouds. If you don't care, you can put everything into one Google Cloud region. You can't use global as your location.

For cluster, we recommend choosing the name of the cluster in which the operator is deployed.

When properly configured, your OperatorConfig should look like this:

    apiVersion: monitoring.googleapis.com/v1
    kind: OperatorConfig
    metadata:
      namespace: gmp-public
      name: config
    collection:
      credentials:
        name: gmp-test-sa
        key: key.json
    rules:
      credentials:
        name: gmp-test-sa
        key: key.json

And your Deployment resource should look like this:

apiVersion: apps/v1
kind: Deployment
...
spec:
  ...
  template:
    ...
    spec:
      ...
      containers:
      - name: operator
        ...
        args:
        - ...
        - "--project-id=PROJECT_ID"
        - "--cluster=CLUSTER_NAME"
        - "--location=REGION"

This example assumes you have set the REGION variable to a value like us-central1, for example.

Running Managed Service for Prometheus outside of Google Cloud incurs data transfer fees. There are fees to transfer data into Google Cloud, and you might incur fees to transfer data out of another cloud. You can minimize these costs by enabling gzip compression over the wire through the OperatorConfig. Add the text shown in bold to the resource:

    apiVersion: monitoring.googleapis.com/v1
    kind: OperatorConfig
    metadata:
      namespace: gmp-public
      name: config
    collection:
      compression: gzip
      ...

Further reading on managed collection custom resources

For reference documentation about all the Managed Service for Prometheus custom resources, see the prometheus-engine/doc/api reference.

What's next

• Use PromQL in Cloud Monitoring to query Prometheus metrics.
• Use Grafana to query Prometheus metrics.
• Use PromQL alerts in Cloud Monitoring.
• Set up managed rule evaluation.
• Set up commonly used exporters.