VMware NSX-T Data Center can be upgraded to 3.2.1 using the upgrade coordinator. The upgrade coordinator resides in the NSX Manager and orchestrates the upgrade process of ESXi hosts, NSX-T Data Center Edge cluster, NSX-T Data Center management plane. The upgrade process involves upgrading the components in the following sequence: VMware NSX-T Data Center Edges, ESXi Hosts, and NSX-T Data Center management plane.

Prerequisites

Follow these prerequisites in NSX-T Data Center:

  1. Perform the Pre-Upgrade Tasks.

  2. Review the NSX-T Data Center Upgrade Checklist to track your work on the NSX-T Data Center upgrade process.

  3. Prepare your infrastructure for the NSX-T Data Center upgrade.

  4. Back up the NSX Manager for 3.1.2 or Back up the NSX Manager for 3.2.0.1

Procedure

  • Upgrade NSX-T Data Center to 3.2.1.
    1. Upgrade the Upgrade Coordinator component.
    2. Upgrade NSX Edge Cluster.
    3. Configure and Upgrade Hosts.
      Note:
      • If ESXi host has VMs with SR-IOV network adapters, then the user needs to be manually power off these VMs for ESXi host to enter maintenance mode.

      • If these VMs are Tanzu Kubernetes Grid worker nodes then drain the pods from these worker nodes, before powering worker node VMs off, for ESXi host to enter maintenance mode. kubectl drain <worker_node_name> --ignore-daemonsets --delete-local-data --force

    4. Upgrade Management Plane.

What to do next

  1. Perform the post-upgrade tasks to verify that the NSX-T Data Center upgrade is successful.

  2. After the NSX-T Data Center upgrade is successful, verify that the clusters prepared for NSX-T Data Center are in the Success state:

    1. Log in to NSX-T manager using Virtual IP (VIP) address.

    2. Navigate to System > Fabric > Node > Host Transport Nodes.

    3. Verify that all the clusters prepared for NSX-T Data Center are in Success state without any error.

    4. If there is an error on the ESXi hosts in the cluster, reconfigure the transport node by selecting the CONFIGURE NSX option on the cluster using the existing transport node profile. If there is any other upgrade issue, see Troubleshooting Upgrade Failures.