---
title: "Capacity planning"
url: https://docs.corefinity.com/guides/corecluster/capacity-planning
type: guide
last_updated: 2026-11-13
feature: capacity-planning
---
# Capacity planning

How a CoreCluster measures its capacity, when it warns and stops new sites, and what to do then.

A CoreCluster's package is one allowance of CPU, memory and storage for all its sites, with no fixed number of sites. CPU and memory count what the sites use; storage counts each site's whole disk.

_Screenshot: The capacity card of a CoreCluster: CPU, memory and storage as shares of the CC8 package, the verdict Fine, and markers at 75% and 90%._

| Any of CPU, memory or storage at | The cluster says | What it means                                                                            |
| -------------------------------- | ---------------- | ---------------------------------------------------------------------------------------- |
| Under 75%                        | **Fine**         | There is room for more sites.                                                            |
| 75%                              | **Upgrade soon** | The cluster is filling up. A new site is still allowed while it fits.                    |
| 90%                              | **Full**         | No new site or environment can be created until the package is bigger or sites use less. |

A new site or environment must fit as well: its expected CPU and memory and its whole disk are added to what the cluster uses now. It is refused, with a message naming what is short, if that takes CPU, memory or storage to 90%, even while the cluster says **Fine** or **Upgrade soon**.

## Check that a site will fit

| Where                                              | What you see                                                                          |
| -------------------------------------------------- | ------------------------------------------------------------------------------------- |
| **Create a managed cluster**, step **First sites** | **Fits on** shows the share of the package the listed sites are expected to use.      |
| **Upgrade package**                                | Each bigger package shows what CPU use would become, such as **CPU now 59 % → 39 %**. |
| A cluster's **Usage**                              | Average and peak use over up to 30 days, against the package line.                    |

## Make room

- Upgrade the package or add capacity by the unit: see [Package, capacity and server type](/guides/corecluster/package-capacity-and-server-type).
- Delete sites and environments you no longer need: see [Delete a site](/guides/sites/create-and-manage/delete-a-site).
- Put new sites on another cluster: see [Create a cluster](/guides/corecluster/create-a-cluster).

## Automate this

**API:** [Check what fits a CoreCluster](https://docs.corefinity.com/api/reference/corecluster/check-cluster-capacity.md), scope `read`

```bash
curl -X POST https://manage.corefinity.com/api/corestack/clusters/1017/capacity/check \
  -H "Authorization: Bearer $COREFINITY_TOKEN" \
  -H "Accept: application/json" \
  -H "Content-Type: application/json" \
  -d '{
    "sites": [
      {
        "platform": "wordpress",
        "tier": "S"
      }
    ]
  }'
```

**MCP:** [`corecluster_capacity_check`](https://docs.corefinity.com/mcp/tools/corecluster_capacity_check.md), scope `read`

```json
{
  "name": "corecluster_capacity_check",
  "arguments": {
    "cluster_id": 812,
    "sites": [
      {
        "platform": "wordpress",
        "tier": "M"
      }
    ]
  }
}
```

**Control Panel:** Managed Clusters › your cluster › Overview ([https://manage.corefinity.com/go/clusters](https://manage.corefinity.com/go/clusters))

## Related

- [Cluster overview](/guides/corecluster/cluster-overview): Check a cluster's capacity, sites and activity.
- [Usage and activity](/guides/corecluster/usage-and-activity): See a cluster's usage and what changed.
- [Package, capacity and server type](/guides/corecluster/package-capacity-and-server-type): Upgrade a cluster's package or add capacity.
