> ## Documentation Index
> Fetch the complete documentation index at: https://docs.e6data.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Kubernetes configuration

> Reference for the Kubernetes resources an In-VPC e6data deployment installs - CRDs, storage class, node provisioning, and namespaces.

This reference summarizes the Kubernetes resources an In-VPC (CloudPrem) deployment installs and the conventions it follows. For the build steps, see the In-VPC deployment guides for [AWS](/platform/deploy/aws-cloudprem/configure-registry-vpc-eks-networking) and [Azure](/platform/deploy/azure-cloudprem/configure-registry-kubernetes-networking).

## Namespaces

| Namespace | Contents |
| - | - |
| `e6operator` | The e6data operator and its CRDs |
| `cert-manager` | cert-manager components |
| `<workspace>` | Per-workspace components (console, query router, metadata services) |

## Custom resource definitions

The e6data operator installs CRDs under `e6data.io`, including `namespaceconfigs`, `queryrouters`, `queryservices`, `metadataservices`, `monitoringservices`, and `e6catalogs`. Clusters are managed as `QueryService` resources - see [CloudPrem cluster operations](/query/clusters/cloudprem-operations).

## Platform components

| Component | Role |
| - | - |
| e6data operator | Reconciles e6data CRDs into running workloads |
| cert-manager | TLS certificate lifecycle |
| Karpenter | Node provisioning via NodePool - EC2NodeClass on AWS, AKSNodeClass on Azure |
| AWS Load Balancer Controller (AWS) | Provisions load balancers for ingress |

## Conventions

* A default **`gp3` StorageClass** (AWS) backs persistent volumes.
* CRDs are applied with **`kubectl apply --server-side`** (the combined payload exceeds Helm's 1 MB release-Secret and 256 KB annotation limits). On AWS the platform and workspace components use the **`helm template | kubectl apply --server-side`** pattern; on Azure the operator is installed with `helm upgrade` and workspace resources are applied as custom resources.
* Workspace pods are scheduled onto workspace node pools via tolerations and node selectors.

## See also

* [CloudPrem cluster operations](/query/clusters/cloudprem-operations)
* [AWS configuration](/platform/reference/config/aws) · [Azure configuration](/platform/reference/config/azure)
* [Helm values](/platform/reference/config/helm-values)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.