Skip to main content
Version: latest

Prometheus Operator

Prometheus is an open-source monitoring system that is designed to collect and analyze metrics from various sources, such as applications, servers, and networks. It is widely used in the DevOps world to monitor the health and performance of applications and infrastructure. Prometheus stores metrics in a time-series database and provides a query language for analyzing the data. It also includes a powerful alerting system that can notify operators when thresholds are breached.

The Prometheus Operator is a tool that simplifies the deployment and management of Prometheus in a Kubernetes cluster. It automates tasks such as configuring Prometheus, creating and managing Prometheus rules and alerts and scaling Prometheus instances based on demand. The Operator uses Kubernetes custom resources to define and manage Prometheus instances and related resources, such as ServiceMonitors, which enable Prometheus to discover and monitor services running in the cluster.

You can use the Prometheus Operator to create a monitoring stack that other host clusters point to and forward metrics to. Check out the guide Deploy Monitoring Stack to learn how to create a monitoring stack with Prometheus for your Palette environment.


We recommend you use version v44.3.x or greater moving forward for a simplified and improved user experience when creating a monitoring stack for your architecture. Starting with version v44.3.x the remote monitoring feature is supported. Check out the Prometheus Remote Write Tuning to learn more about the remote monitoring feature.

Versions Supported


  • Kubernetes v1.16 or greater.

  • The minimum required size for the Prometheus server is 4 CPU, 8 GB Memory, and 10 GB Storage. We recommend the monitoring stack have 1.5x to 2x the minimum required size:

    Recommended size:

    • 8 CPU
    • 16 GB Memory
    • 20 GB Storage

    As new clusters with the Prometheus agent are added to your environment, review the resource utilization and consider increasing resources if needed. As the Prometheus documentation recommends, each additional agent requires the following resources from the monitoring stack:

    Each added agent:

    • 0.1 CPU
    • 250 MiB Memory
    • 1 GB Storage

    Refer to the Prometheus Operational aspects documentation for additional guidance.


The Prometheus operator supports all the parameters exposed by the kube-prometheus-stack Helm Chart. Refer to the kube-prometheus-stack documentation for details.

The Prometheus Operator pack has one parameter you must initialize grafana.adminPassword:

adminPassword: ""

Use the grafana.adminPassword parameter to assign a password to the Grafana admin user admin.

Additional parameters you should be aware of can be found by expanding the Presets options. You can modify the preset settings when you create the profile or when you deploy the cluster and review the cluster profile.

A view of the pack's preset drawer expanded with radio buttons

Review the usage section below to learn more about each preset option.


Check out the guide Deploy Monitoring Stack to learn how to create a monitoring stack with Prometheus for your Palette environment.

Airgap Palette and VerteX

In the context of an airgap Palette or VerteX installation, you must remove the grafana-piechart-panel plugin. This plugin is not included by default and requires an internet connection to download. The monitoring stack doesn't require this plugin to function properly.

- grafana-piechart-panel

Email Alerts

You can configure the Prometheus server to send email alerts to a set of contacts. Toggle the Email Alerts button to enable email alerting. Update the alertmanager.config.receivers settings with all the required email setting values.

- name: email-alert
- to: <reciever>@<domainname>.com
send_resolved: true
from: <sender>@<domainname>.com
smarthost: smtp.<domainname>.com:587
auth_username: <sender>@<domainname>.com
auth_identity: <sender>@<domainname>.com
auth_password: <sender_passwd>

Refer to the Prometheus Alertmanager Configuration documentation to learn more about Alertmanager.

Grafana Ingress

You can enable an ingress endpoint for Grafana that will deploy an NGINX ingress controller. This feature can be used to enable HTTPS and require authentication for all Prometheus API requests.

If you do not enable the ingress option, then by default a service with a load balancer will be created that exposes port 80.

Toggle the Enable button to enable the use of Ingress.

Thanos SideCar

Thanos is an open-source system for running large-scale, distributed, and highly available Prometheus setups. Thanos allows Prometheus to store data for extended periods in object storage, such as Amazon S3 or Google Cloud Storage, instead of a local disk. This enables Prometheus to scale horizontally without the risk of using up local storage space.

Toggle the Enable button to enable the use of Thanos.

Object Store

Select the Thanos object storage type you will use. Review the thanos.objstoreConfig parameters to configure the use of object storage with Thanos. Refer to the Thanos Object Storage documentation to learn more about how to configure each object storage.


Thanos Ruler Object Store

By default, Thanos Ruler event data is saved in object storage specified for Thanos, but you can specify a different object storage for event data. Refer to the Thanos Ruler resource to learn more.

Remote Monitoring

You can configure the Prometheus server to accept metrics from Prometheus agents and become a centralized aggregation point for all Kubernetes metrics. Enabling this feature will expose port 9090 of the prometheus-operator-prometheus service. Use the generated service URL to provide other Kubernetes clusters with the installed Prometheus Agent so that cluster metrics can be forwarded to the Prometheus server.

The remote monitoring feature is configured with defaults to help you consume this feature out-of-the-box. You can change any configuration related to remote monitoring to fine-tune settings for your environment.

Refer to the Prometheus Remote Write Tuning resource to learn more about configuration options.

To get started with remote monitoring, check out the Deploy Monitoring Stack guide.

Palette Resources Monitoring

You can access internal Palette metrics in Grafana by adding the Prometheus Cluster Metrics pack to all your client clusters. Refer to the Enable Monitoring on Host Cluster guide to learn more.

Persistent Storage

You can configure the Prometheus Operator to use persistent storage. To enable persistent storage add the following code snippet to the kube-prometheus-stack.prometheus.prometheusSpec.storageSpec configuration block in the pack's YAML configuration file. The code snippet below creates a Persistent Volume Claim (PVC) for the Prometheus Operator.

name: prom-operator-pvc
storageClassName: spectro-storage-class
accessModes: ["ReadWriteOnce"]
storage: 5Gi


The Prometheus Operator pack installs the following dependencies:


You can retrieve details about the Prometheus operator pack by using the following Terraform code.

data "spectrocloud_registry" "public_registry" {
name = "Public Repo"

data "spectrocloud_pack_simple" "pack-info" {
name = "prometheus-opeartor"
version = "45.4.0"
type = "helm"
registry_uid =