Deploy NSX Edge Cluster without Tier-0/1 in VMware Cloud Foundation 5.1

Recently, I needed to deploy an NSX Edge Cluster Without T0 and T1.

The reason was due to VCD integration the T1s aren’t managed directly on NSX but on VCD side.
Another reason can also be that the exposition of the T0; the eBGP peering may be initiated on another broadcast domain than the Overlay.

This advanced configuration is a brand-new feature of VCF 5.1:

Tier-0/1 optional for VCF Edge cluster:
When creating an Edge cluster with the VCF API, the Tier-0 and Tier-1 gateways are now optional.

https://docs.vmware.com/en/VMware-Cloud-Foundation/5.1/rn/vmware-cloud-foundation-51-release-notes/index.html*

Deploy NSX Edge Cluster without Tier-0/1 in VMware Cloud Foundation 5.1
is only available through API.

For doing this advanced configuration, you can use your usual API coding solution but do not forget to first log yourself and reuse your bearer token.

But there is an easier way: usually, I use the “Developer Center” of VCF for pulling my configuration.

This is the code that you need to type:

{

    "edgeRootPassword": "TypeYourRootPassword",
    "edgeAdminPassword": "TypeYourAdminPassword",
    "edgeAuditPassword": "TypeYourAuditPassword",
    "edgeClusterName": "NameOfCluster",
    "edgeClusterProfileType": "DEFAULT",
    "edgeClusterType": "NSX-T",
    "edgeFormFactor": "LARGE",
    "mtu": 9000,
    "edgeNodeSpecs": [
        {
            "clusterId": "IDofTheCluster",
            "edgeNodeName": "FQDNOfTheNode",
			"edgeTepGateway" : "GWOfTheNode",
			"edgeTep1IP" : "NodeTepIP1InCIDR",
			"edgeTep2IP" : "NodeTepIP2InCIDR",
            "edgeTepVlan": VlanID,
            "interRackCluster": false,
            "managementGateway": "MngGWIP",
            "managementIP": "MngIPInCIDR"
        },
		{
            "clusterId": "IDofTheCluster",
            "edgeNodeName": "FQDNOfTheNode",
			"edgeTepGateway" : "GWOfTheNode",
			"edgeTep1IP" : "NodeTepIP1InCIDR",
			"edgeTep2IP" : "NodeTepIP2InCIDR",
            "edgeTepVlan": VlanID,
            "interRackCluster": false,
            "managementGateway": "MngGWIP",
            "managementIP": "MngIPInCIDR"
        }
    ],
	"skipTepRoutabilityCheck": false
}

To grab the clusterId, you need to run the query API GET /v1/clusters and find the cluster “ID” attribute below the cluster name where you want to deploy your edge cluster.

Remark 1, you do not need to specify the management VLAN or Portgroup, VCF will reuse the one defined in the Network Pool of the Domain.

Remark 2, you can deploy cluster across cluster/domain(vCenter) by changing the ClusterID

Before applying the code, test it with the API call: POST /v1/edge-clusters/validations

To apply the code you need to use the API call POST /v1/edge-clusters

Source of my code: here

If the task failed due to wrong input you can:

1. Remove the task by typing in ssh this command :
   curl -X DELETE 127.0.0.1/tasks/registrations/IDOfTheTask
   • IDOfTheTask can be grabbed by clicking on the task and looking at the address bar
2. Use the remove Edge Cluster script provided by Broadcom
   • KB 316110

Have a nice reading,

Cheers

Valentin