Create Cluster
Pre-requisites to a successful cluster creation
Before creating clusters, Atmosly validates the following on your cloud account:
AWS (EKS)
- The account has not exceeded the VPC limit
- The account can have KMS keys created
- EIP quota on the account is within limits (minimum 5 available)
- Instance limit has not been reached
- IAM roles limit has not been exceeded
GCP (GKE)
- The project has sufficient quota for the selected region/zone
- Required APIs are enabled (Kubernetes Engine API, Compute Engine API)
- IAM permissions are properly configured
Atmosly checks for the above while creating clusters.
Create a cluster
- Navigate to the Cluster module from the left menu.
- Click on Add Cluster inside Cluster actions in the top right corner.
- Next, fill in the required fields.
info
The Target account will populate based on the cloud integrations (AWS or GCP) you have configured.
When selecting the VPC option, you can choose either an existing VPC or create a new one based on your requirements.
- In the cluster type section, choose which type of environment you want to create — prod or non-prod.
- Based on your selection, you will need a few additional details for a successful cluster creation.
- Select an AMI (AWS) or image type (GCP) for the cluster nodes.
- Select the add-ons you want to have on your cluster, preview the configuration and the estimated cost, and CREATE!
tip
You can customize the configuration of add-ons like PGL Stack, ArgoFlow, K8s Dashboard, and others.
Alternatively, you also have the option to create clusters in ONE CLICK. Navigate to One-Click Cluster to learn more.
Cluster Configuration Fields
Refer to the table below for various fields and values associated with creating clusters:
| Name | Description |
|---|---|
| Display Name (mandatory) | What you want your cluster to be called. Name cannot contain special characters and should be unique. |
| Target Account (mandatory) | The cloud account (AWS or GCP) against which you want your cluster created. |
| Region (mandatory) | Cloud provider region in which you want your cluster. For GCP, you may also select a specific zone. |
| Environment (mandatory) | Select the type of environment — prod or non-prod (Staging, QA, etc). |
| Select VPC | You can either choose an existing VPC or create a new one. (AWS only — GCP uses VPC networks from your project.) |
| AMI / Image Type | AWS: Select from AMI families — AL2 (Amazon Linux 2), AL2023 (Amazon Linux 2023), Ubuntu, Bottlerocket, Windows2019, Windows2022. GCP: Select image type — COS_CONTAINERD, COS, UBUNTU, UBUNTU_CONTAINERD, WINDOWS_SAC, WINDOWS_LTSC. |
| Availability Zones (mandatory) | Once the region is selected, you will be able to see available AZs. Each region has multiple Availability Zones for redundancy. |
| Kubernetes Version (mandatory) | Atmosly supports Kubernetes versions 1.24 through 1.33. Each version includes end-of-standard-support and end-of-extended-support dates for tracking. |
| VPC CIDR (mandatory, AWS) | The IP address range for your VPC. Default: 10.0.0.0/16. |
| Tags | Key-value pairs for resource identification, organization, and billing tracking. |
| Cluster and VPC Flow Log (mandatory for prod) | Cluster Flow Log: Captures IP traffic information for your cluster's network interfaces. VPC Flow Log: Captures IP traffic going to and from network interfaces in your VPC. |
| Cluster Log Retention Period | Duration for which cluster logs are kept. Default: 90 days. |
| VPC Flow Log Retention Period | Duration for which VPC flow logs are stored. Default: 60 days. |
| VPC Flow Log Max Aggregation Interval | Frequency for publishing flow log records to CloudWatch. Default: 600 seconds. |
| Control Plane Logging | Sends EKS/GKE control plane audit and diagnostic logs to your cloud provider's logging service. |
It can take up to 20 minutes for a cluster to become ACTIVE on Atmosly.
Add-ons
Atmosly provides 23 add-ons across multiple categories that can be installed on your cluster:
Networking & Ingress
| Add-on | Description | Cloud |
|---|---|---|
| NGINX Ingress Controller | Handles routing of external HTTP/S traffic to Kubernetes services | All |
| Ingress ALB Controller | Manages AWS Application Load Balancer instances for routing external traffic | AWS |
| VPC CNI | AWS VPC Container Network Interface plugin for pod networking | AWS |
| Istio Service Mesh | Service mesh for traffic management, security, and observability | All |
Certificates & Security
| Add-on | Description | Cloud |
|---|---|---|
| Cert Manager | Automates management and issuance of TLS certificates | All |
| Cert Manager LE Email | Let's Encrypt email configuration for Cert Manager | All |
| Letsencrypt HTTP Issuers | HTTP-01 challenge issuers for Let's Encrypt certificates | All |
| External Secrets | Syncs secrets from external secret management systems (AWS Secrets Manager, GCP Secret Manager, etc.) | All |
Observability & Monitoring
| Add-on | Description | Cloud |
|---|---|---|
| PGL Stack | Prometheus for monitoring, Grafana for visualization, and Loki for log aggregation | All |
| K8s Dashboard | Web interface for managing and monitoring Kubernetes clusters | All |
| Metrics Server | Resource metrics pipeline for Kubernetes (CPU/memory usage) | All |
| Service Monitor CRD | Custom Resource Definitions for Prometheus service monitoring | All |
| HPA for CoreDNS | Horizontal Pod Autoscaler configuration for CoreDNS | All |
Auto-scaling
| Add-on | Description | Cloud |
|---|---|---|
| Karpenter | Just-in-time node provisioning for Kubernetes workloads | AWS |
| Cluster Autoscaler | Automatically adjusts cluster size based on resource demands | All |
| Cluster Proportional Autoscaler | Scales workloads proportional to cluster size | All |
| Keda | Event-driven autoscaling for Kubernetes workloads | All |
Storage
| Add-on | Description | Cloud |
|---|---|---|
| EFS Storage Class | Integrates AWS Elastic File System with Kubernetes for persistent storage | AWS |
| EBS CSI Driver | AWS Elastic Block Store Container Storage Interface driver | AWS |
| Single AZ GP3 SC | Single Availability Zone GP3 storage class | AWS |
CI/CD & Workflow
| Add-on | Description | Cloud |
|---|---|---|
| ArgoFlow | Enables Argo Workflows for CI engine capability on the cluster | All |
Node Management
| Add-on | Description | Cloud |
|---|---|---|
| AWS Node Termination Handler | Gracefully handles EC2 instance interruptions (spot termination, scheduled maintenance) | AWS |
| Reloader | Automatically restarts pods when ConfigMaps or Secrets change | All |
note
Some add-ons are cloud-specific (AWS or GCP). Add-on availability is filtered based on your cluster's cloud provider and Kubernetes version. Add-ons may have dependencies on other add-ons, which are automatically resolved during installation.