Home / CMDB / Hosts, Clusters & Datacentres
CMDB

Hosts, Clusters & Datacentres

Hosts are the physical hypervisor servers (ESXi, AHV, Hyper-V) that run your virtual machines. Clusters group hosts together into resource pools. Datacentres represent physical facilities. These three asset types form the infrastructure foundation for capacity planning and Nutanix Move migrations.

On this page

Overview

In a virtualised infrastructure migration, knowing where virtual machines live — and where they are going — is as important as knowing what those VMs do. The Hosts, Clusters, and Datacentres module captures this physical and logical infrastructure topology.

Hosts are physical hypervisor servers: VMware ESXi nodes, Nutanix AHV hosts, or Hyper-V servers. Each host record tracks its total CPU, memory, and which cluster it belongs to. The VMs (Devices) running on each host are visible from the Host Detail page.

Clusters aggregate the resources of all member hosts and expose total CPU, memory, and storage capacity. The Capacity Planning dashboards pull from Cluster records to calculate how much headroom exists on the target infrastructure for incoming migrated workloads.

Datacentres represent physical facilities. They provide geographic and compliance context for capacity planning and SLA reporting — for example, distinguishing workloads that must remain on-premise in a specific region from those being migrated to cloud.

In most projects, host, cluster, and datacentre records are auto-populated by the VMware or Nutanix integration. Manual entry is only needed for environments not reachable by those integrations.

When to use this

  • Pre-migration capacity planning: Verify that target cluster resources are sufficient before assigning move groups.
  • Source infrastructure discovery: Run the VMware or Nutanix integration to populate host and cluster records automatically.
  • Manual inventory entry: Add hosts manually for isolated or air-gapped environments not reachable by the discovery integrations.
  • Datacentre mapping: Register source and target datacentres to provide geographic context in dashboards and reports.
  • Host-to-VM relationship review: Use the Host Detail page to see which VMs are running on a specific host — useful when planning host-level maintenance windows.

Key features

Host Inventory
Register and manage all physical hypervisor servers. View hosted VMs, resource allocation, and cluster membership per host.
Cluster Resource Aggregation
Cluster records aggregate CPU, memory, and storage totals from member hosts. These figures drive the Capacity Planning dashboards.
Datacentre Registry
Track physical facilities with location, country, and Uptime Institute tier rating. Supports multi-site and hybrid migration programmes.
VMware & Nutanix Integration
Auto-populate hosts and clusters via the VMware vCenter or Nutanix Prism integrations. No manual entry required for reachable environments.
Capacity Planning Feed
Host and cluster data feeds directly into the Capacity Planning module, showing available headroom on target infrastructure.
Infrastructure Topology View
Drill down from Datacentre → Cluster → Host → VM to visualise your full physical infrastructure hierarchy.

Step-by-step instructions

Running VMware or Nutanix discovery (recommended)

1
Go to Integrations
Navigate to Integrations → VMware / RVTools or Integrations → Nutanix depending on your hypervisor platform.
2
Configure credentials
Enter the vCenter or Prism Central address, username, and password (stored securely in Credentials). Test the connection before proceeding.
3
Run the discovery
Click Run Discovery. Clarity Migrate will enumerate all hosts, clusters, VMs, and resource pools. This typically takes 1–5 minutes depending on estate size.
4
Review and confirm imported records
After discovery completes, navigate to CMDB → Hosts and CMDB → Clusters to review the imported records. Check resource totals for accuracy.

Adding a host manually

1
Navigate to Hosts
Go to CMDB → Hosts and click + Add Host in the toolbar.
2
Enter host details

Complete the form with:

  • Hostname — the physical server name
  • IPv4 / IPv6 Address
  • CPU Count and Memory (GB)
  • Hypervisor Type — ESXi, AHV, or Hyper-V
  • Cluster — the cluster this host belongs to (create the cluster first if needed)
3
Save
Click Save. The host now appears in the Host List and its resources are aggregated into the linked cluster's totals.

Adding a cluster

1
Navigate to Clusters
Go to CMDB → Clusters and click + Add Cluster.
2
Enter cluster details
Enter the Name, Type (VMware vSphere, Nutanix AHV, Hyper-V), and Datacentre. CPU, memory, and storage totals will be auto-calculated from linked hosts once they are added.
3
Link hosts to the cluster
After saving, either add hosts and assign them to this cluster, or edit existing host records to update their Cluster field. The cluster's resource totals will update automatically.

Adding a datacentre

1
Navigate to Datacentres
Go to CMDB → Datacentres and click + Add Datacentre.
2
Enter the facility details
Enter the facility Name, Location (city/address), Country, and Tier (Uptime Institute Tier I–IV). Mark whether it is a source or target datacentre for the migration.
3
Save and link clusters
Click Save. Edit cluster records to assign them to this datacentre so the full hierarchy is established.

Important fields

Host fields

Field Required / Optional Description
Hostname Required The physical server hostname. Should match the name visible in vCenter, Prism, or Hyper-V Manager.
IPv4 / IPv6 Optional Management IP address of the hypervisor host. Used for network mapping and discovery deduplication.
CPU Count Optional Total physical CPU cores on the host. Aggregated to cluster totals and used in capacity planning calculations.
Memory (GB) Optional Total physical RAM in GB. Aggregated to cluster totals for capacity headroom analysis.
Hypervisor Type Required The virtualisation platform: VMware ESXi, Nutanix AHV, or Microsoft Hyper-V. Required for tool compatibility checks.
Cluster Optional The cluster this host belongs to. Required for resource aggregation in capacity planning. Hosts without a cluster are excluded from cluster-level calculations.

Cluster fields

Field Required / Optional Description
Name Required The cluster name as seen in vCenter or Prism Central.
Type Required Cluster technology: VMware vSphere, Nutanix AHV, or Hyper-V Failover Cluster.
Datacentre Optional The physical facility this cluster resides in. Used to link clusters to datacentre records for geographic reporting.
Node Count Optional Number of host nodes in the cluster. Auto-calculated when hosts are linked.
Total CPU / Memory / Storage Optional Aggregate resource totals. Auto-calculated from linked host records when populated. Used directly in Capacity Planning dashboards.

Datacentre fields

Field Required / Optional Description
Name Required The common name for this facility (e.g., London DC1, Azure UK South).
Location Optional City or address of the physical facility.
Country Optional Country of the facility. Used for data sovereignty reporting and compliance filtering.
Tier Optional Uptime Institute Tier rating (I–IV). Tier IV is the highest availability rating. Used in SLA and resilience reporting.

Example workflow

Real-world scenario
Verifying target cluster capacity before wave planning begins

Before assigning move groups for Wave 1, an infrastructure engineer needs to confirm that the target Nutanix cluster has sufficient CPU and memory headroom to absorb the incoming workloads.

She navigates to Integrations → Nutanix and runs a discovery against the target Prism Central instance. After 3 minutes, the discovery completes and the Cluster List now shows the target Nutanix cluster — NTX-PROD-01 — with 384 CPU cores and 3,072 GB RAM available.

She then navigates to Migrations → Capacity Planning and selects the NTX-PROD-01 cluster. The dashboard shows current allocation (from VMs already running on the cluster) and projects the additional load from the 40 devices in Wave 1's move group. With 30% headroom remaining after Wave 1, she confirms the cluster can absorb the workload and gives the wave plan a green light.

She also checks the Cluster List to confirm that all 8 AHV hosts in NTX-PROD-01 are listed and have accurate CPU/memory values — two were added manually without memory figures, which she corrects by editing the host records.

Tips & best practices

Run discovery first — don't add hosts manually unless required

The VMware and Nutanix integrations populate hosts, clusters, VMs, and resource data far more accurately and completely than manual entry. Always run discovery first. Manual entry should only be used for isolated environments that the integrations cannot reach.

Cluster resource totals power capacity planning

The Capacity Planning module reads CPU, memory, and storage totals directly from Cluster records. If cluster totals are missing or incorrect — either because hosts aren't linked or CPU/memory fields are blank — the capacity calculations will be wrong. Verify cluster resource totals before beginning wave assignment.

Register both source and target datacentres

Create Datacentre records for both your source and target environments. This enables the migration dashboard to show workloads by facility and lets you track progress at a datacentre level — useful for programmes migrating out of a specific facility.

Common mistakes

Creating duplicate host records

If discovery has already run and you then manually add the same host, you'll have duplicate records. This inflates cluster resource totals and creates confusion in the Host List. Always search for an existing host record before adding manually. If duplicates already exist, use the Data Quality screen to identify and merge them.

Not setting the Cluster field on host records

A host without a Cluster assignment is not included in any cluster's resource totals. This means the Capacity Planning dashboard will undercount available capacity and may incorrectly indicate that a cluster is at risk of oversubscription. Always set the Cluster field on every host record.

Forgetting to re-run discovery after infrastructure changes

If new hosts are added to your environment after the initial discovery, they won't appear in Clarity Migrate automatically. Re-run the VMware or Nutanix discovery periodically — especially before each wave's capacity check — to ensure the CMDB reflects the current state of your infrastructure.