Enabling vSphere Workload Management

User Guides for VMware Solutions
Audience
Public
Content Type
User Guides
Source Type
Documentation

This section will show an example of how to enable Workload Management using Everpure FlashArray. Before proceeding please make sure to confirm that all other steps listed in the Prerequisites section above have been completed or Workload Management will not work properly.

To get started, click on the main vSphere client menu and select the Workload Management option. Note that this option is only available in vSphere 7 and up.

Click on the Clusters tab and select Add Cluster.

In the window that opens one or more compatible clusters will be shown where Workload Management can be enabled. Select the desired Cluster and click Next.

Depending on the use case, select the appropriate Control Plane size. Smaller exploratory test/dev instances should use the Tiny or Small options while larger production kubernetes deployments should consider using Medium or Large to have the appropriate amount of resources available. Once the Size is selected click Next.

The next couple steps in Workload Management enablement are assigning the correct networking components. We will break out each entry in the networking section individually here:

  1. Network: Select the VM Management Network
  2. Starting IP Address: Enter the starting IP address that resides on the Management Network selected in the previous step. Note that 5 consecutive IP addresses must be available. So in this example, 10.21.143.225 - 10.21.143.229 are all free for use by the Supervisor Cluster.
  3. Subnet Mask for Management Network.
  4. Gateway for Management Network.
  5. DNS Server IP or FQDN. While optional having this field populated is recommended.
  6. NTP Server: It is recommended to set this field to whatever you are using for your ESXi hosts and vCenter instance to ensure all components remain in time sync.
  7. Optionally include any DNS Search Domains.

    Here are the relevant entries for the 2nd section of Networking Information:

  8. Select the Distributed vSwitch that your Edge Cluster resides on.
  9. Select the NSX-T Edge Cluster you wish to use.
  10. Optionally enter the API Server Endpoint FQDN for Kubernetes. This entry needs to be pre-populated in DNS and will be associated to the Ingress CIDR entry below in field 7 below. For this example the API Server endpoint FQDN would have an IP address of 10.21.132.65 set in DNS.
  11. Enter a DNS server that can be reached by the VDS selected in the first step.
  12. Pod CIDRs: No need to change this from default unless that network range is already in use.
  13. Service CIDRs: No need to change this from default unless that network range is already in use.
  14. Ingress CIDRs: This needs to be a minimum of a /27 network to be usable. The Ingress CIDR also needs to be on the same network as one of the Uplinks that were defined on the Tier0 Gateway from the Edge Node VM.
  15. Egress CIDRs: This needs to be a minimum of a /27 network to be usable. The Ingress CIDR also needs to be on the same network as one of the Uplinks that were defined on the Tier0 Gateway from the Edge Node VM. Note that in this example we are using a single VLAN for both ingress and egress.

    With all of the above fields populated, click on Next to proceed.

    The Supervisor Control Plane has multiple storage components associated with it. In this section of Workload Management is where we select the SPBM policy previously created earlier within this user guide. Both vVols or VMFS are supported for the Supervisor cluster.

    Click on the Select Storage text.

    From the Select Storage Policies window, pick a VMFS-based SPBM policy and click OK.

    Repeat this procedure for Ephemeral Disks and Image Cache.

    With a policy now associated with all 3 storage pieces of the Control Plane, click on Next.

    In the summary window you have a chance to review how Workload Management will be deployed. Click on Finish to begin enabling Workload Management and note that it can take anywhere from 30 minutes to an hour to complete.

    The relevant log file for Workload Management can be viewed from an ssh session to vCenter via running the following command:

    
tail -f /var/log/vmware/wcp/wcpsvc.log

    Here we can see that Workload Management has been successfully enabled and that the API endpoint for Kubernetes can be found on an IP address specified on our Ingress CIDR.

    Workload Management Enablement Technical Video Demo