> For the complete documentation index, see [llms.txt](https://docs.silk.us/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.silk.us/deploy/plan-your-deployment/cluster-and-datapod-sizing.md).

# Cluster and DataPod Sizing

Understand how cluster size, deployment profile, and node configuration determine the scale, performance, and capacity of your Silk environment.

Sizing a Silk environment involves two related decisions:

* **Cluster sizing** defines the maximum number of DataPods and nodes that a Flex environment can support.
* **DataPod sizing** defines the performance, storage capacity, and scaling characteristics available to an individual DataPod.

Choose a cluster that can handle your long-term environment growth, then configure each DataPod based on its workload, performance, capacity, and cloud infrastructure needs.

{% hint style="warning" %}
**Plan for future growth**

You can't change the cluster size once the cluster is created. Increasing or decreasing the cluster size requires replacing the existing cluster. Plan for expected DataPod and node growth before deployment.
{% endhint %}

## Choose a cluster size

A Flex instance manages one cluster, which can contain multiple DataPods. The cluster size establishes the maximum number of DataPods, c.nodes, and m.nodes the environment can support.

Each cluster size also requires a specific IP address range.

| Cluster Size | Maximum DataPods                                                               | Maximum m.nodes                                                                | Maximum c.nodes                                                                                     | Required IP range |
| ------------ | ------------------------------------------------------------------------------ | ------------------------------------------------------------------------------ | --------------------------------------------------------------------------------------------------- | ----------------- |
| **Small**    | 2                                                                              | <p><strong>AWS</strong>: 2<br><strong>Azure and Google Cloud</strong>: 3</p>   | <p><strong>AWS and Azure</strong>: 9<br><strong>Google Cloud</strong>: 10</p>                       | `/24`             |
| **Medium**   | 4                                                                              | <p><strong>AWS</strong>: 4<br><strong>Azure and Google Cloud</strong>: 6</p>   | <p><strong>AWS and Azure</strong>: 23<br><strong>Google Cloud</strong>: 24</p>                      | `/23`             |
| **Large**    | <p><strong>AWS</strong>: 17<br><strong>Azure and Google Cloud:</strong> 24</p> | <p><strong>AWS</strong>: 17<br><strong>Azure and Google Cloud</strong>: 25</p> | <p><strong>AWS</strong>: 98<br><strong>Azure</strong>: 99<br><strong>Google Cloud</strong>: 100</p> | `/21`             |

Select a cluster size based on the combined requirements of all DataPods you expect to deploy.

Consider:

* The number of DataPods required for workload, application, or organizational separation
* The number of c.nodes and m.nodes each DataPod may require
* Planned capacity or performance expansion
* DataPods that may be added for development, testing, disaster recovery, or other future use cases
* The address space your organization can reserve for the Silk environment

A larger cluster establishes a higher scaling limit but doesn't provision all the supported DataPods or nodes at deployment. Cloud resources are added as DataPods are created and expanded.

## Size a DataPod

A DataPod combines compute and storage resources to provide block storage services to applications and databases. Its configuration includes three node types:

* **c.nodes** provide the compute resources used to process I/O and deliver data services.
* **m.nodes** organize storage resources into scalable capacity layers and provide data durability to c.nodes.
* **d.nodes** provide the underlying persistent storage and are grouped to form m.nodes.

DataPod sizing isn't down to a single capacity selection. It combines a workload profile, storage architecture, VM types, node counts, and an allocated capacity limit.

{% stepper %}
{% step %}

### Select a workload profile

The workload profile establishes the available node options, capacity limits, and scaling characteristics for the DataPod.

<table><thead><tr><th width="244.073974609375">Workload profile</th><th>Best suited for</th></tr></thead><tbody><tr><td><strong>High Performance</strong></td><td>Most production workloads that require balanced performance and scalability. This is the default profile.</td></tr><tr><td><strong>Accelerated Performance</strong></td><td>Workloads that require higher sustained throughput and lower latency. This profile uses read cache and may require additional cloud resources or quota.</td></tr><tr><td><strong>Limited Capacity</strong></td><td>Smaller-scale workloads that require less logical capacity. Availability depends on the cloud provider and deployment architecture.</td></tr></tbody></table>

Available profiles and configuration options vary by cloud provider and storage type. See the [DataPod Deployment Profiles](/learn/products-and-core-concepts/datapod-architecture/datapod-deployment-profiles.md) documentation to learn more.
{% endstep %}

{% step %}

### Size for performance with c.nodes

c.nodes provide the DataPod's processing resources. Increasing the number of c.nodes can increase supported throughput and IOPS, subject to the limits of the selected cloud VM type and cloud provider.

Standard DataPods require at least two c.nodes. AWS supports up to seven c.nodes per DataPod, while Azure and Google Cloud support up to eight c.nodes per DataPod.

When sizing c.nodes, consider:

* Required read and write throughput
* IOPS and latency requirements
* Expected concurrent workload activity
* Cloud VM bandwidth and sustained-throughput limits
* Available requirements during maintenance or recovery
* Capacity requirements associated with the selected profile

A DataPod must retain the minimum number of online c.nodes required for its configuration. Operations that remove a c.node require that minimum plus one additional c.node while recovery is in progress.

{% hint style="info" %}
We recommend using the default c.node VM type presented in Flex. Alternate types can help address temporary resource availability constraints, but they may have different performance characteristics or quota requirements.
{% endhint %}
{% endstep %}

{% step %}

### Size for storage with m.nodes

Every DataPod requires at least one m.node. The maximum supported number depends on the storage architecture:

* **VM Local Disk deployments** support up to four m.nodes.
* **Azure Managed Disk deployments** support up to three m.nodes.

m.node sizes and supported VM families vary by cloud provider. In supported deployments, you can select multiple compatible m.node VM types within the same DataPod to balance capacity, performance, resource availability, and cloud cost.

For VM Local Disk deployments, the selected m.node VM determines the underlying d.node configuration and physical capacity. For Azure Managed Disk deployments, the configured Premium SSD v2 resources determine the available capacity.

{% hint style="warning" %}
You can't mix VM Local Disk and Managed Disk m.nodes within the same DataPod, except temporarily during supported m.node live migration operations.
{% endhint %}
{% endstep %}

{% step %}

### Set the capacity limit

Physical and logical capacity represent different aspects of DataPod sizing:

* **Physical capacity** is the underlying storage available from the DataPod's storage resources.
* **Logical capacity** is the maximum capacity that can be allocated to volumes and other logical objects before data reduction.

Silk measures capacity in tebibytes (TiB), where 1 TiB equals approximately 1.10 TB.

For High Performance and Accelerated Performance deployments, the supported capacity limits depend on the initial c.node configuration:

| Minimum c.nodes | Minimum online c.nodes | Maximum physical capacity | Maximum logical capacity |
| --------------- | ---------------------- | ------------------------- | ------------------------ |
| 2               | 1                      | 324 TiB                   | 512 TiB                  |
| 3               | 2                      | 360 TiB                   | 1,024 TiB                |

Limited Capacity deployments provide smaller logical capacity increments.

The options available during DataPod creation depend on the cloud provider and storage architecture.
{% endstep %}
{% endstepper %}

## Evaluate cloud infrastructure

Before finalizing the configuration, confirm that the selected region and subscription/project can provide the required resources.

Review:

* Supported VM families and instance types
* CPU, memory, and storage quotas
* Managed disk quotas, when applicable
* Instance and storage availability in the selected region or zone
* Network address space required by the cluster
* VM-level bandwidth and sustained-performance limits
* Any additional resources required by the selected deployment profile

## Review sizing with Silk

Work with the Silk Cloud Infrastructure team to complete the Silk Checklist before initial cluster and DataPod deployment. The checklist validates that your selected cloud provider, region, deployment profile, and node configuration align with your workload and performance requirements.

It also confirms that the proposed physical and logical capacity, cloud quotas, and resource availability can support both your current environment and anticipated growth.

After deployment, use Flex and the SDP Management Console to monitor capacity and performance. You can scale supported DataPods by adding or removing resources within the limits established by the cluster and the DataPod's initial configuration.

## Related resources

{% content-ref url="/pages/FXudgU7rJdcdoOSoDNH5" %}
[Plan DataPod Capacity](/deploy/plan-your-deployment/plan-datapod-capacity.md)
{% endcontent-ref %}

{% content-ref url="/spaces/jI1rjszhUJU6LG8J53d0/pages/jbwHJE9XlrTdPS1lXZgj" %}
[Clusters and DataPods](/learn/products-and-core-concepts/datapod-architecture/clusters-and-datapods.md)
{% endcontent-ref %}

{% content-ref url="/spaces/jI1rjszhUJU6LG8J53d0/pages/RACFGQWRzeDLCRkQE8tg" %}
[DataPod Deployment Profiles](/learn/products-and-core-concepts/datapod-architecture/datapod-deployment-profiles.md)
{% endcontent-ref %}


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.silk.us/deploy/plan-your-deployment/cluster-and-datapod-sizing.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
