theVirtualPaddy

Applying Transport Node Profile is not allowed when vLCM Config Manager is enabled on the cluster

—

by

Jameseydoyle
in

I came across an issue with a customer earlier today that impacts the installation of NSX 4.1 on a vSphere 8.0 cluster with the new image-based vLCM configured.

The error message you might see when you try to apply a transport node profile to a cluster is as follows:
Error: vLCM Config Manager is enabled on this cluster. Applying Transport Node Profile is not allowed. (Error code: 26219)

Normally, you should be able to use a Transport Node Profile to configure a cluster with vLCM enabled using a single image, so what is the problem here?
https://docs.vmware.com/en/VMware-NSX/4.1/installation/GUID-0E4F8397-AD9E-40C8-9B71-B2CEF53A2AD4.html

The issue here is that the cluster is also configured to use a single configuration, which is set at the time of the cluster creation.

If this option is enabled on a cluster, it prevents NSX from configuring each host individually, which it needs to do to install NSX on the cluster, resulting in the error shown above.

To workaround this issue, you must create a new cluster with the option disabled, and move your hosts from the old cluster into the new cluster. Then you will be able to apply the Transport Node Profile to the new cluster.

Comments

6 responses to “Applying Transport Node Profile is not allowed when vLCM Config Manager is enabled on the cluster”

  1. fireballg Avatar
    fireballg

    But how am I supposed to remove the hosts from a cluster if it’s vSAN-enabled?

    Like

    Reply
    1. Jameseydoyle Avatar
      Jameseydoyle

      You would need to put the host into maintenance mode, remove the disk groups, and then remove the host from the cluster. Once out of the cluster, you could remove the vSAN VMkernel ports, or reimage the host.

      Like

      Reply
      1. Jonas Nagel Avatar
        Jonas Nagel

        Maybe I was unclear, but I’ve got 3 hosts with vSphere 8.0U3 which are in a vSAN cluster and I want to apply an NSX TP to it, which doesn’t work because it’s already managed by vLCM. Now, I can’t simply remove them, because it would destroy the environment. Do I have to manually install the VIB/ZIP on each host? And if yes, does that still work after all?

        Like

      2. Jameseydoyle Avatar
        Jameseydoyle

        That might work, but I suspect you might still get the same error when you try to configure NSX at the cluster level. I’d need to lab it, but I don’t have access to a lab at the moment.

        You haven’t specified whether you are using vSAN ESA or OSA, but the procedure would basically be the same. As you only have three hosts, this procedure will require that you run the risk of running with reduced redundancy for a time. You would create a new cluster, take one host out of the existing cluster (wiping the vSAN configuration and disk groups first), and put it in the new cluster with a default storage policy with Force Provisioning turned on. Force provisioning will allow vSAN to create objects even if there are not enough resources to satisfy the storage policy. You will then be able to vMotion VMs into the new cluster where new storage objects will be created with reduced redundancy (i.e. one storage component and no mirrors or witness components). Once the VMs have been migrated, you can then remove the second and third hosts from the original cluster and introduce them to the new cluster (you should remove the disk groups and wipe the disks first to ensure you do not end up with a lot of orphaned components which would consume space). When all three hosts are in the new cluster, you can remove the Force Provisioning option from the storage policy, which will prompt the creation of the redundant components for the objects.

        I would suggest opening a ticket with VMware Support if you have a support contract to help guide you through this process.

        Like

      3. Jameseydoyle Avatar
        Jameseydoyle

        Your main challenge is that you only have the bare minimum number of hosts to support a vSAN cluster. Ideally, you should use 4 hosts to give you the extra capacity you need to deal with long term hardware failures or issues like this.

        Like

      4. fireballg Avatar
        fireballg

        Thanks for your reply; it’s a lab environment which was setup before I realized I could not enable it for NSX, so don’t worry too much about long term issues.

        Like

Leave a reply to fireballg Cancel reply