OneLens Installation Scripts

Simplified Kubernetes cost optimization and monitoring deployment

Overview

OneLens deploys a monitoring stack into your Kubernetes cluster to collect cost and resource utilization data. The deployment consists of two parts:

OneLens Deployer (this chart) - A Kubernetes Job that installs and configures the monitoring stack. A daily CronJob keeps it updated.
OneLens Agent (installed by the deployer) - The monitoring stack: OneLens Agent, Prometheus, OpenCost, and Kube-State-Metrics.

You only install the deployer chart. It handles everything else.

Prerequisites

Kubernetes cluster (1.25+)
Helm 3.0+
kubectl configured for your cluster
AWS EBS CSI driver (for AWS EKS clusters) or Azure Disk CSI driver (for AKS clusters)
Minimum node resources available: 50m CPU and 256Mi memory for the deployer job

Run the pre-requisite checker to validate your environment before installing. It checks connectivity, tools, Kubernetes version, and CSI driver status:

curl -sSL https://raw.githubusercontent.com/astuto-ai/onelens-installation-scripts/master/scripts/prereq-check/onelens-prereq-check.sh | bash

Quick Start

1. Add the Helm repository

OneLens charts are hosted on a public Helm repository. Add it to your local Helm client so you can install charts from it:

helm repo add onelens https://astuto-ai.github.io/onelens-installation-scripts/
helm repo update

2. Install

This installs the OneLens deployer, which registers your cluster and sets up the full monitoring stack. You only need to do this once per cluster — running it again on an already-connected cluster will fail at registration.

Your Kubernetes clusters are automatically discovered and visible in the OneLens console. Navigate to the cluster you want to connect, and the console provides a ready-to-use install command with the REGISTRATION_TOKEN pre-filled. Copy and run it directly, or use the template below:

helm upgrade --install onelensdeployer onelens/onelensdeployer \
  -n onelens-agent --create-namespace \
  --set job.env.CLUSTER_NAME=<cluster-name> \
  --set job.env.REGION=<region> \
  --set-string job.env.ACCOUNT=<account-id> \
  --set job.env.REGISTRATION_TOKEN=<token>

Add any optional parameters (encryption, labels, tolerations) from the Configuration Reference below.

3. Verify installation

# Check all pods are running
kubectl get pods -n onelens-agent

# Expected pods (all should be Running):
#   onelens-agent-prometheus-server-*        - Metrics storage
#   onelens-agent-kube-state-metrics-*       - Kubernetes object metrics
#   onelens-agent-prometheus-opencost-*      - Cost metrics
#   onelens-agent-prometheus-pushgateway-*   - Metrics push endpoint
#
# Note: The onelens-agent pod is a CronJob that runs hourly by default.
# It collects metrics from Prometheus and sends them to the OneLens API.
# It will not appear until its first scheduled run. To trigger it immediately,
# see step 4 below.

4. Trigger a manual data collection (optional)

kubectl create job manual-trigger --from=cronjob/onelens-agent -n onelens-agent

Configuration Reference

All parameters below are passed via --set flags during helm upgrade --install. Examples are shown with each section so you can copy-paste and adapt.

Required Parameters — cluster name, region, account, token
Storage Encryption — encrypt Prometheus persistent volumes (AWS EBS / Azure Disk)
Volume Tags — apply custom tags to persistent volumes for cost tracking
Node Scheduling — run OneLens pods on dedicated or specific nodes
Labels — apply custom labels to all OneLens resources
Other — image pull secrets, CronJob schedule, suspend updater

Required Parameters

Parameter	Description
`job.env.CLUSTER_NAME`	Your Kubernetes cluster name
`job.env.REGION`	Cloud region (e.g., `us-east-1`, `centralindia`)
`job.env.ACCOUNT`	Cloud account ID (use `--set-string` to preserve leading zeros)
`job.env.REGISTRATION_TOKEN`	Registration token from OneLens platform

AWS EKS example

helm upgrade --install onelensdeployer onelens/onelensdeployer \
  -n onelens-agent --create-namespace \
  --set job.env.CLUSTER_NAME=my-eks-cluster \
  --set job.env.REGION=us-east-1 \
  --set-string job.env.ACCOUNT=123456789012 \
  --set job.env.REGISTRATION_TOKEN=your-token

Azure AKS example

helm upgrade --install onelensdeployer onelens/onelensdeployer \
  -n onelens-agent --create-namespace \
  --set job.env.CLUSTER_NAME=my-aks-cluster \
  --set job.env.REGION=centralindia \
  --set-string job.env.ACCOUNT=your-subscription-id \
  --set job.env.REGISTRATION_TOKEN=your-token

Storage Encryption

OneLens creates a StorageClass for Prometheus persistent volumes. You can enable encryption on these volumes.

AWS EBS

Parameter	Description	Default
`job.env.EBS_ENCRYPTION_ENABLED`	Enable EBS volume encryption	`false`
`job.env.EBS_ENCRYPTION_KEY`	Custom KMS key ARN (omit to use AWS default `aws/ebs` key)	`""`

Encrypt with the default AWS-managed key (aws/ebs):

helm upgrade --install onelensdeployer onelens/onelensdeployer \
  -n onelens-agent --create-namespace \
  --set job.env.CLUSTER_NAME=my-eks-cluster \
  --set job.env.REGION=us-east-1 \
  --set-string job.env.ACCOUNT=123456789012 \
  --set job.env.REGISTRATION_TOKEN=your-token \
  --set job.env.EBS_ENCRYPTION_ENABLED=true

Encrypt with a customer-managed KMS key:

helm upgrade --install onelensdeployer onelens/onelensdeployer \
  -n onelens-agent --create-namespace \
  --set job.env.CLUSTER_NAME=my-eks-cluster \
  --set job.env.REGION=us-east-1 \
  --set-string job.env.ACCOUNT=123456789012 \
  --set job.env.REGISTRATION_TOKEN=your-token \
  --set job.env.EBS_ENCRYPTION_ENABLED=true \
  --set job.env.EBS_ENCRYPTION_KEY=arn:aws:kms:us-east-1:123456789012:key/your-key-id

Azure Disk

Parameter	Description	Default
`job.env.AZURE_DISK_ENCRYPTION_ENABLED`	Enable Azure Disk encryption	`false`
`job.env.AZURE_DISK_ENCRYPTION_SET_ID`	Azure Disk Encryption Set resource ID	`""`
`job.env.AZURE_DISK_CACHING_MODE`	Disk caching mode (`None`, `ReadOnly`, `ReadWrite`)	`ReadOnly`

helm upgrade --install onelensdeployer onelens/onelensdeployer \
  -n onelens-agent --create-namespace \
  --set job.env.CLUSTER_NAME=my-aks-cluster \
  --set job.env.REGION=centralindia \
  --set-string job.env.ACCOUNT=your-subscription-id \
  --set job.env.REGISTRATION_TOKEN=your-token \
  --set job.env.AZURE_DISK_ENCRYPTION_ENABLED=true \
  --set job.env.AZURE_DISK_ENCRYPTION_SET_ID=/subscriptions/<sub-id>/resourceGroups/<rg>/providers/Microsoft.Compute/diskEncryptionSets/<des-name>

Volume Tags

Apply custom tags to the persistent volumes created by OneLens. Useful for cost tracking and compliance.

AWS EBS

Parameter	Description	Default
`job.env.EBS_TAGS_ENABLED`	Enable custom tags on EBS volumes	`false`
`job.env.EBS_TAGS`	Comma-separated `key=value` pairs	`""`

  --set job.env.EBS_TAGS_ENABLED=true \
  --set job.env.EBS_TAGS="env=prod,team=platform,cost-center=engineering"

Azure Disk

Parameter	Description	Default
`job.env.AZURE_DISK_TAGS_ENABLED`	Enable custom tags on Azure Disks	`false`
`job.env.AZURE_DISK_TAGS`	Comma-separated `key=value` pairs	`""`

  --set job.env.AZURE_DISK_TAGS_ENABLED=true \
  --set job.env.AZURE_DISK_TAGS="env=prod,team=platform,cost-center=engineering"

Node Scheduling

Schedule OneLens pods on specific nodes using nodeSelector and tolerations. The job.env.* parameters apply to the agent pods (Prometheus, KSM, OpenCost, etc.). To also schedule the deployer job/cronjob on the same nodes, set job.tolerations, job.nodeSelector, cronjob.tolerations, and cronjob.nodeSelector as shown in the example.

Parameter	Description	Default
`job.env.NODE_SELECTOR_KEY`	Node selector label key (applied to all agent pods)	`""`
`job.env.NODE_SELECTOR_VALUE`	Node selector label value	`""`
`job.env.TOLERATION_KEY`	Toleration key (applied to all agent pods)	`""`
`job.env.TOLERATION_VALUE`	Toleration value (leave empty for `Exists` operator)	`""`
`job.env.TOLERATION_OPERATOR`	`Equal` or `Exists`	`""`
`job.env.TOLERATION_EFFECT`	`NoSchedule`, `PreferNoSchedule`, or `NoExecute`	`""`

Example — nodes tainted with dedicated=onelens:NoSchedule:

helm upgrade --install onelensdeployer onelens/onelensdeployer \
  -n onelens-agent --create-namespace \
  --set job.env.CLUSTER_NAME=my-cluster \
  --set job.env.REGION=us-east-1 \
  --set-string job.env.ACCOUNT=123456789012 \
  --set job.env.REGISTRATION_TOKEN=your-token \
  --set job.env.NODE_SELECTOR_KEY=dedicated \
  --set job.env.NODE_SELECTOR_VALUE=onelens \
  --set job.env.TOLERATION_KEY=dedicated \
  --set job.env.TOLERATION_VALUE=onelens \
  --set job.env.TOLERATION_OPERATOR=Equal \
  --set job.env.TOLERATION_EFFECT=NoSchedule \
  --set job.nodeSelector.dedicated=onelens \
  --set 'job.tolerations[0].key=dedicated' \
  --set 'job.tolerations[0].operator=Equal' \
  --set 'job.tolerations[0].value=onelens' \
  --set 'job.tolerations[0].effect=NoSchedule' \
  --set cronjob.nodeSelector.dedicated=onelens \
  --set 'cronjob.tolerations[0].key=dedicated' \
  --set 'cronjob.tolerations[0].operator=Equal' \
  --set 'cronjob.tolerations[0].value=onelens' \
  --set 'cronjob.tolerations[0].effect=NoSchedule'

For taints without a value (e.g., dedicated:NoSchedule), use Exists operator and omit the value:

  --set job.env.TOLERATION_OPERATOR=Exists \
  --set-string job.env.TOLERATION_VALUE="" \
  --set 'job.tolerations[0].operator=Exists'

Labels

Apply custom labels to OneLens resources. Useful for organizational policies that require specific labels on all resources.

Parameter	Description	Default
`globals.labels`	Applied to namespace, deployer job/cronjob, and all agent pods	`{}`
`job.labels`	Additional labels only on the deployer job	`{}`
`cronjob.labels`	Additional labels only on the updater cronjob	`{}`

  --set globals.labels."company\.com/team"=platform \
  --set globals.labels."company\.com/env"=prod

Other

Parameter	Description	Default
`job.env.IMAGE_PULL_SECRET`	Image pull secret name for private registries	`""`
`cronjob.schedule`	Updater CronJob schedule	`"0 2 * * *"`
`cronjob.suspend`	Suspend the daily updater	`false`

Upgrade

To upgrade to a newer version:

helm repo update
helm upgrade onelensdeployer onelens/onelensdeployer \
  -n onelens-agent \
  --version <new-version> \
  --reuse-values

The deployer job will re-run and upgrade the agent stack to the matching version.

Uninstall

# Remove the deployer
helm uninstall onelensdeployer -n onelens-agent

# Remove the agent stack (installed by the deployer)
helm uninstall onelens-agent -n onelens-agent

# Optionally delete the namespace and all resources
kubectl delete namespace onelens-agent

Note: PersistentVolumeClaims are retained by default (helm.sh/resource-policy: keep). To also delete Prometheus data:

kubectl delete pvc -n onelens-agent --all

Troubleshooting

See the Troubleshooting Guide for common issues, diagnostic commands, and operational procedures.

How It Works

helm install onelensdeployer
    |
    v
[Deployer Job] -- runs install.sh inside a pod
    |              - detects cloud provider (AWS/Azure)
    |              - registers cluster with OneLens API
    |              - sizes resources based on pod count
    |              - creates StorageClass, RBAC, namespace
    |              - runs: helm install onelens-agent
    v
[OneLens Agent Stack]
    - Prometheus (metrics collection + storage)
    - Kube-State-Metrics (Kubernetes object metrics)
    - OpenCost (cost calculation)
    - OneLens Agent (data processing + upload to OneLens platform)
    - Pushgateway (metrics push endpoint)
    |
    v
[Daily Updater CronJob] -- runs patching.sh
    - re-evaluates cluster size
    - adjusts resource allocations
    - applies configuration updates

Documentation

CI/CD Architecture - Complete CI/CD pipeline documentation
Quick Reference - Fast commands and troubleshooting
Release Process - How to create new releases
Configuration Guide - Detailed configuration options

Scripts & Tools

Pre-requisite Checker - Validate your environment before installation
EBS Driver Installation - Install AWS EBS CSI driver with IAM roles
Dedicated Node Setup - Create tainted node pools for OneLens

Support

Email: support@astuto.ai
Documentation: OneLens Docs
Issues: GitHub Issues

Name		Name	Last commit message	Last commit date
Latest commit History 623 Commits
.github		.github
charts		charts
docs		docs
lib		lib
scripts		scripts
src		src
tests		tests
.gitignore		.gitignore
Dockerfile		Dockerfile
README.md		README.md
VERSION		VERSION
entrypoint.sh		entrypoint.sh
globalvalues.yaml		globalvalues.yaml
install.sh		install.sh
patching.sh		patching.sh

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Repository files navigation

OneLens Installation Scripts

Table of Contents

Overview

Prerequisites

Quick Start

1. Add the Helm repository

2. Install

3. Verify installation

4. Trigger a manual data collection (optional)

Configuration Reference

Required Parameters

Storage Encryption

Volume Tags

Node Scheduling

Labels

Other

Upgrade

Uninstall

Troubleshooting

How It Works

Documentation

Scripts & Tools

Support

About

Uh oh!

Releases 16

Contributors

Uh oh!

Languages

Folders and files

Latest commit

History

Repository files navigation

OneLens Installation Scripts

Table of Contents

Overview

Prerequisites

Quick Start

1. Add the Helm repository

2. Install

3. Verify installation

4. Trigger a manual data collection (optional)

Configuration Reference

Required Parameters

Storage Encryption

Volume Tags

Node Scheduling

Labels

Other

Upgrade

Uninstall

Troubleshooting

How It Works

Documentation

Scripts & Tools

Support

About

Resources

Uh oh!

Stars

Watchers

Forks

Releases 16

Contributors

Uh oh!

Languages