If you have both NSX-T and Enterprise PKS workload domains in your environment, Enterprise PKS workload domains must be upgraded before NSX-T domains.
Upgrade Enterprise PKS Workload Domains
Within an Enterprise PKS workload domain, the following components need to be upgraded. They can be upgraded in any order.
- vCenter Server and Platform Services Controllers. See Upgrade vCenter Server and Platform Services Controllers.
- ESXi. See Upgrade ESXi.
- Enterprise PKS. See Upgrade Enterprise PKS to VMware Cloud Foundation 3.10.1.2.
Upgrade NSX-T Workload Domains
Within an NSX-T workload domain, components must be upgraded in the following order.
- If you have both Enterprise PKS and NSX-T workload domains in your environment, you must upgrade Enterprise PKS workload domains before NSX-T domains.
- NSX-T. See Upgrade NSX-T Data Center.
- vCenter Server and Platform Services Controllers. See Upgrade vCenter Server and Platform Services Controllers.
- ESXi. See Upgrade ESXi.
Upgrade NSX for vSphere Workload Domains
If you performed a skip-level upgrade, you may also need to upgrade NSX for vSphere workload domains.
- NSX for vSphere. See Upgrade NSX for vSphere.
- vCenter Server and Platform Services Controllers. See Upgrade vCenter Server and Platform Services Controllers.
- ESXi. See Upgrade ESXi.
Post Upgrade Steps for NFS-Based Workload Domains
After upgrading workload domains, you must add a static route for hosts to access NFS storage over the NFS gateway. This process must be completed before expanding the workload domain.
- Identify the IP address NFS server for the workload domain.
- Identify the network pool associated with the hosts in the cluster and the NFS gateway for the network pool.
- Log in to SDDC Manager.
- Click and then click the workload domain you are performing the post upgrade steps on.
- Click the Clusters tab and then click an NFS-based cluster.
- Click the Hosts tab and note down the network pool for the hosts.
- Click the Info icon next to the network pool name and note down the NFS gateway.
- Ensure that the NFS server is reachable from the NFS gateway. If a gateway does not exist, create it.
- Identify the vmknic on each host in the cluster that is configured for NFS traffic.
- Configure a static route on each host to reach the NFS server from the NFS gateway.
esxcli network ip route ipv4 add -g NFS-gateway-IP -n NFS-gateway
- Verify that the new route is added to the host using the NFS vmknic
esxcli network ip route ipv4 list
- Ensure that the hosts in the NFS cluster (nfs-cluster-1) can reach the NFS gateway (10.0.24.1) through the NFS vmkernel (vmk2) using the vmkping command.
vmkping -4 -I vmk2 -s 1470 -d -W 5 10.0.22.250
- Repeat steps 2 through 7 for each cluster using NFS storage.