Pairing more configuration flexibility with more integrated automation delivers even more simplified outcomes to meet more business needs!

More is what sums up this latest Cloud Foundation on VxRail release! This new release is based on the latest software bill of materials (BOM) featuring vSphere 8.0 U2, vSAN 8.0 U2, and NSX 4.1.2. Read on for more details.…

Operations and serviceability user experience updates

SDDC Manager WFO UI custom host networking configuration enhancements

With this enhancement, the administrator can configure networking of a new workload domain or VxRail cluster using either “Default” VxRail Network Profiles or a “Custom” Network Profile configuration. Cloud Foundation on VxRail already supports the ability for administrators to deploy custom host networking configurations using the SDDC Manager WFO API deployment method, however this new feature now brings this support to the SDDC Manager WFO UI deployment method, making it even easier to operationalize.

The following demo walks through using the SDDC Manager WFO UI to create a new workload domain with a VxRail cluster that is configured with vSAN ESA and VxRail vLCM mode enabled and a custom network profile.

New VCF Infrastructure as Code (IaC) tooling with new Terraform VCF Provider and PowerCLI VCF Module

Infrastructure teams can now utilize the Terraform Provider for VCF and the VCF module that is now integrated into VMware’s official PowerCLI tool to perform Infrastructure-as-code (IaC), allowing them to deploy, manage, and operate VMware Cloud Foundation on VxRail deployments. 

By using prebuilt IaC best practices code that is designed to take advantage of interfacing with a single VCF API, IaC teams are able to perform infrastructure provisioning tasks that can accelerate IaC usage and lessen the burden to develop and maintain code for individual infrastructure components intended to deliver similar outcomes.

Important Note: Not all operations using these tools may be supported in Cloud Foundation on VxRail. Please refer to tool documentation links at the bottom of this post for details.

LCM updates

Day 1 VxRail vLCM mode compatibility for management and workload domains

VMware Cloud Foundation 5.1 on VxRail 8.0.200 now supports the configuration and deployment of new domains using vSphere Lifecycle Manager Images (vLCM) enabled VxRail clusters, depicted in figure 1. VxRail vLCM enabled clusters can leverage VxRail Manager to unify not only your ESXi Image but also your BIOS/firmware/drivers through a single update process, all controlled/orchestrated by VxRail Manager using the integrated SDDC Manager’s native LCM operations experience via VxRail APIs. VxRail clusters will have their VxRail Continuously Validated State image managed at the cluster level by VxRail Manager just like in VxRail standard LCM mode enabled clusters. 

Figure 1. High-level VxRail vLCM mode architecture

Mixed-mode support for workload domains as a steady state

Existing VMware Cloud Foundation 5.x on VxRail 8.x deployments now allow administrators to run workload domains of different VCF 5.x versions as a “steady state”.  Administrators can now update the management domain and any other workload domain of a VCF 5.0 deployment to the latest VCF 5.x version without the need to upgrade all workload domains. Mixed-mode support also allows administrators to leverage the benefits of new SDDC Manager features in the management domain without having to upgrade a full VCF 5.x on VxRail 8.x instance. 

Asynchronous download support for SDDC Manager update precheck files

SDDC Manager update precheck files can now be downloaded and updated asynchronously from full release updates, an addition to similar async VxRail specific precheck file updates that already exist within VxRail Manager. This feature allows administrators to download, deploy, and run SDDC Manager update prechecks tailored to a specific VMware Cloud Foundation on VxRail releases. SDDC Manager precheck files are created by VMware engineering and contain detailed checks for SDDC Manager to run prior to upgrading to a newer VCF on VxRail target release, as shown in the following figure. 

Figure 2. High-level process of asynchronous download support for SDDC Manager update precheck files

Networking updates

Support for the separation of DvPG for management appliances and ESXi host (VMKernel) management

Prior to this release, the default networking topology deployed by VMware Cloud Foundation on VxRail consisted of ESXi host management interfaces (vmkernel interface) and management components (vCenter server, SDDC Manager, NSX components, VxRail Manager, etc.) being applied to the same Distributed Virtual Port Group (DvPG). This new DvPG separation feature enables traffic isolation between management component VMs and ESXi Host Management vmkernel Interfaces, helping align to an organization’s desired security posture. Figure 3 illustrates this new configuration architecture.

Figure 3. New DvPG architecture

Configure custom NSX Edge cluster without 2-tier routing (via API)

VMware Cloud Foundation 5.1 on VxRail 8.0.200 now provides the option to deploy a custom NSX Edge cluster without the need to configure both a Tier-0 and Tier-1 gateway. These types of NSX Edge cluster deployments can be configured using the SDDC Manager (API only).

Static IP-based NSX Tunnel End Point and Sub Transport Node Profile assignment support for L3 aware clusters and L2/L3 vSAN stretched clusters

VxRail stretched clusters that are deployed using vSAN OSA can now be configured with vLCM mode enabled. In addition, administrators can now configure NSX Host TEPs to utilize a NSX static IP pool and no longer need to manually maintain an external DHCP server to support Layer 3 vSAN OSA stretched clusters, as illustrated in the following figure.

Figure 4. TEP Configuration Flexibility Example for vSAN Stretched Clusters

Building off these capabilities, deployments of VxRail stretched clusters with vSAN OSA which are configured using static IP Pools can now also leverage Sub-Transport Node Profiles (Sub-TNP), a feature introduced with NSX-T 3.2.2 and NSX 4.1.

Sub-TNPs can be used to prepare clusters of hosts without L2 adjacency to the Host TEP VLAN. This is useful for customers with rack-based IP schemas and allows Host TEP IPs to be configured on their own separate networks. Configuring vSAN stretched clusters using NSX Sub-TNP provides increased security, allowing administrators to enable and configure Distributed Malware Prevention and Detection. An example of this is depicted in the following figure.

Figure 5. Sub-TNP vSAN L3 Stretched Cluster Configuration Example

Note: Stretched VxRail with vSAN ESA clusters are not yet supported.

Support for multiple VDS for NSX host networking configurations

This release now provides the option to configure multiple VDS for NSX through the SDDC Manager WFO UI and WFO API.

Administrators can now configure additional VxRail host VDS prepared for NSX (VDS for NSX) to configure using VLAN Transport Zones (VLAN TZs), as shown in the following figure. This provides administrators the added benefit of configuring NSX Distributed Firewall (DFW) for workloads in VLAN transport zones, allowing security to be more granular. These capabilities further simplify the configuration of advanced networking and security for Cloud Foundation on VxRail.

Figure 6 is VxRail host

Figure 6. Configuring additional VxRail host VDS for NSX to configure using VLAN TZs

Security and access updates

OKTA SSO identity federation support

VMware Cloud Foundation 5.1 on VxRail 8.0.200 now supports the option to configure the VMware Identity Broker for federation using Okta (3rd party IDP).  Once configured, federated users can seamlessly move between vCenter Server and NSX Manager consoles without being prompted to re-authenticate.  

Storage updates

vSAN OSA/ESA support for management and workload domain VxRail clusters

VMware Cloud Foundation 5.1 on VxRail 8.0.200 adds support for both vSAN OSA-based and vSAN ESA-based VxRail clusters when deploying a new management domain (greenfield VCF on VxRail instance) and new workload domains/clusters in VCF on VxRail instances that have been upgraded to this latest release. VCF requires that vSAN ESA-based cluster deployments have vLCM mode enabled. Also, as of this release, only 15^th generation VxRail vSAN ESA compatible hardware platforms are supported. 16^th generation VxRail platform support is planned for a future release.

Support for vSAN OSA/ESA remote datastores as principal storage when used with VxRail dynamic node workload domain clusters

This release adds support of VxRail dynamic node compute-only clusters in cross cluster capacity sharing use cases. This means that vSAN OSA or ESA remote datastores sourced from a standard VxRail HCI cluster with vSAN within the same workload domain can now be used as principal storage for VxRail dynamic node- compute only workload domain clusters. This capability is available via the SDDC Manager WFO script deployment method only. 

Platform and scale updates

Increased VCF remote cluster maximum support for up to 16 nodes and up to 150ms latency

There are new validated updates to the maximum supported latency requirements for use of VCF remote clusters. These links now require 10 Mbps of bandwidth available and a latency less than 150ms.

There have also been updates regarding VCF remote cluster size scalability ranges. A VCF remote cluster now requires a minimum of 3 hosts when using local vSAN as cluster principal storage or 2 hosts when using supported Dell external storage principal storage with VxRail dynamic nodes. On the max scale limit side, VCF remote clusters cannot exceed the new maximum of 16 VxRail hosts in either case. 

Note: Support for this feature is expected to be available after GA.

Support for 2-node workload domain VxRail dynamic node clusters when using VMFS on FC Dell external storage as principal storage

Cloud Foundation on VxRail now supports the ability to deploy 2-node dynamic node-based workload domain clusters when using VMFS on FC Dell external storage as cluster Principal storage.

Increased GPU scale for Private AI

Nvidia GPUs can be configured for AI / ML to support a variety of different use cases. In VMware Cloud Foundation 5.1 on VxRail 8.0.200, where GPUs have been configured for vGPUs, a VM can now be configured with up to 16 vGPU profiles that represent all of a GPU or parts of a GPU. These enhancements allow customers to support larger Generative AI and large-language model (LLM) workloads while delivering maximum performance.

VxRail hardware platform updates

15^th generation VxRail E660N and P670N all-NVMe vSAN ESA hardware platform support

Cloud Foundation on VxRail administrators can now use VxRail hardware platforms that have been qualified to run vSAN ESA and VxRail 8.0.200 software. The all-NVMe VxRail platforms such as the 15^th generation VxRail E660N and P670N can now be ordered and deployed in Cloud Foundation 5.1 on VxRail 8.0.200 environments.

Hybrid cloud management updates

VCF mixed licensing mode support

VMware Cloud Foundation 5.1 on VxRail 8.0.200 introduces support for both Key-based and Keyless licensing for existing deployments, as illustrated in the following figure. 

To enable the deployment, the management domain must first be cloud connected and subscribed.  Once complete, enhanced SDDC Manager workflows allow administrators the option to license a new workload domain using Keyless licenses (cloud connected subscription) or Key-based licenses (perpetual or cloud disconnected subscription). This deployment scenario is referred to as Mixed Licensing Mode. All licensing used within a domain must be homogenous, meaning all components within a domain must use either a Key-based or Keyless license and not a combination thereof.

Figure 7. Understanding Key-based and Keyless licensing for existing deployments

VMware Cloud Disaster Recovery service for VCF cloud connected subscription deployments

VMware Cloud Foundation on VxRail cloud connected subscriptions now support VMware Cloud Disaster Recovery (VCDR) as an add-on service through the VMware Cloud Portal.

Other asynchronous release-independent related updates

VMware redefines Cloud Foundation product lifecycle policies

The product lifecycle policies for new and existing VMware Cloud Foundation releases have been redefined by VMware. VCF on VxRail product lifecycle policies align with VMware’s VCF product lifecycle policy. 

End of General Support for VCF 5.x is now four (4) years from the original VCF 5.0 launch date. This change allows IT teams to run their VMware Cloud Foundation on VxRail deployments for longer before planning an upgrade, providing more control for IT organizations to adopt a cloud operating model that evolves at the pace of their business. 

Summary

Well, there you have it! Another release in the books. If you want even more information beyond what was discussed here, feel free to check out the resources linked below. See you next time!

Resources

• VxRail product page
• VxRail Info Hub page
• VxRail Videos
• VMware Cloud Foundation on Dell VxRail Release Notes
• VCF on VxRail Interactive Demo
• VMware Product Lifecycle Matrix
• Terraform Provider for VCF
• PowerCLI VCF Module

Author: Jason Marques

Twitter:  @vWhipperSnapper