What’s New in vSphere 8 – Part 1

In the latest VMware Explore US 2022 (previously called VMworld) event, VMware announced the new vSphere 8. vSphere 8 is officially expected to be available by October 28, 2022

Workloads are growing everywhere, and multi-cloud will soon become the primary model for infrastructure implementation. VMware proved that with the launch of vSphere+ and vSAN+, they were built to work on that implementation model(Multi-Cloud) by adding many Cloud management features and options.

Table of Contents

1. What is new in vSphere 8?
2. vSphere Distributed Services Engine
3. What is Data Processing Units (DPU)
4. vSphere with Tanzu
5. Lifecycle Management
6. Enhanced Recovery of vCenter
7. Conclusion

There are many new features and enhancements in this new vSphere 8, and I can’t go through them all, but I will try to focus on the most important ones.

Since presenting all the vSphere 8 news in one article would be a huge article and not easy to read, we will divide it into two parts. This is “What’s New in vSphere 8 – Part 1”

What is new in vSphere 8?

Until now there is no information, that much has changed regarding the maximum compute resources. Maybe until the official release date, something may change in the next build.

As we can see above, vGPU has 8 vGPUs per VM – a 2x increase compared to vSphere 7. vLCM increases from 400 hosts to 1000 hosts that can be managed and 10.000 VMs per Cluster. These last 2 are for sure because vSphere+ will manage many vCenters and Clusters through the new vSphere+ online features.

Regarding Passthrough, it increased from 8 to 32 NVIDIA GPUs devices in Passthrough mode – a 4x increase compared to vSphere 7.

In vSphere 8, GPU and vGPU are a big part of the new features included in this new vSphere version.

I will divide the new features per section:

• vSphere Distributed Services Engine
• vSphere with Tanzu
• Lifecycle Management
• Enhanced Recovery of vCenter

vSphere Distributed Services Engine

Basically, the vSphere Distributed Services Engine enables the offloading of network services to Data Processing Units (DPU), also known as SmartNICs(in the vSphere platform) or Distributed Services Cards (DSCs).

What is Data Processing Units (DPU)

DPU manages devices on that layer, like a NIC or GPU, in the hardware layer. ESXI computer layer runs and manages networking, storage, and host management services.

In vSphere 8, an additional instance of ESXi is installed directly on the DPU. This allows ESXi to offload some ESXi services to the DPU for increased performance.

The lifecycle management of the vSphere Distributed Services Engine will be managed by vLCM. It means that ESXi installation on a DPU can be easily managed through vLCM

vSphere 8 and vSphere Distributed Services Engine includes a Distributed Switch version 8.0 using DPU.

With the new virtual Switch version vDS 8.0, ESXi increases the network performance, security encryption, isolation, and protection you would expect from NSX.

This can be done by using Network Offload Compatibility when selecting the DPU compatibility group option, offering a simple Configuration for Network Offloads.

vSphere with Tanzu

What is new with In vSphere 8 with Tanzu? There are substantial changes on vSphere with Tanzu, and where VMware did most of its changes on vSphere 8.

• Tanzu Kubernetes Grid
• Workload Availability Zones
• ClusterClass
• Flexible Package Management using Tanzu CLI
• Federated Authentication with Pinniped

Tanzu Kubernetes Grid 2.0 new version is the first and one of the most significant changes. It now offers Tanzu Kubernetes Grid into a single unified Kubernetes.

Workload Availability Zones allow Supervisor Clusters and Tanzu Kubernetes Clusters to span across vSphere Clusters for increased availability.

Tanzu Kubernetes Grid Service can now be highly available across multiple zones to increase availability.

ClusterClass is an OpenSource Specification that is part of the ClusterAPI Project. Since it is a default pre-install package, you can create your ClusterClass, reuse and redeploy as many times you want.

ClusterAPI defines a declarative way to lifecycle manage Kubernetes Clusters through an existing Management Kubernetes Cluster. In vSphere with Tanzu, that management cluster is the Supervisor Cluster.

It allows upstream Kubernetes conformant ClusterAPI and defines configuration and default installed packages for Tanzu Kubernetes Clusters.

After a Cluster is deployed, it is possible to add additional packages from the Tanzu Standard Package Repository using Tanzu CLI. This will make life easier for Developers and DevOps.

Include Contour for Ingress to the cluster, Certificate management, Logging, Observability with Prometheus/Grafana or External DNS. These are managed as add-ons through the Tanzu CLI interface.

Finally, in vSphere 8 with Tanzu is Federated Authentication with Pinniped

In the previous version of vSphere with Tanzu, authentication was through vCenter SSO. In vSphere 8, you can still use integration with vCenter SSO, but with Pinniped integration, the Supervisor Cluster and Tanzu Kubernetes Clusters can have direct access to OIDC or LDAP IDP without relying on vCenter Single Sign-On.

Supervisor and TKG Clusters can now support and login integration using Pinniped based authentication through Tanzu CLI.

As we can see above, vSphere with Tazu and Tanzu Kubernetes Grid 2.0 has many changes and features included in the new vSphere 8.

Lifecycle Management

It is also another area where there are many changes and improvements, particularly to be integrated with vSphere+ and also the new DPU support.

• DPU support
• Staging Support
• Parallel Remediation
• vSphere Configuration Profiles
• Standalone Host Support (API only)
• VCG Listings for HSM feature support

Note: Before showing the new features and changes on vSphere Lifecycle Manager(vLCM), be aware that vSphere 8 will be the last release that supports baseline lifecycle management.

Only vSphere Lifecycle Manager (vLCM) images will be supported in the future, and vSphere Update Manager baselines will be deprecated.

As stated at the beginning of this article, Lifecycle Management vSphere 8 introduces DPU support for vSphere Lifecycle Manager(vLCM). vLCM automatically remediates the ESXi installation on a DPU in lock-step with the host ESXi version.

Staging of update/upgrade payloads, parallel remediation, and also standalone host support combine to bring vLCM up to feature parity with the Update Manager.

We can now stage an ESXI host without putting the ESXi host in maintenance mode, reducing the impact that lifecycle operations may have on running VMs and workloads.

All vSphere administrators who do ESXi host remediation know that it is time-consuming with considerable wait times.

Now in vSphere 8 with Parallel Remediation in vSphere Lifecycle Manager, we can remediate multiple hosts in parallel and reduce the time needed to remediate an entire ESXi Cluster.

vSphere Configuration Profiles will replace Host Profiles in the future. vSphere 8 introduces the next generation of cluster configuration management vSphere Configuration Profiles.

The desired configuration is defined at the cluster object applied to all hosts in the cluster. All hosts in the cluster have a consistent configuration applied.

Since vSphere 8 is yet to be launched, some of the new features are still under development. This is the case of vSphere Configuration Profile. VMware states that future releases of vSphere 8 will expand and enhance its support.

But for now, Host Profiles continue to be supported in vSphere 8.

Enhanced Recovery of vCenter

Personally, this is one of the most valuable features in vSphere 8 for vCenter.

Everyone who previously restored a vCenter from a backup knows that vCenter and ESXi hosts show issues, like out-of-sync and Virtual Switches issues after a restore. Particularly if you add an ESXi host or make any changes after a backup and then you do a restore.

In vSphere 8 with Distributed Key-Value Store (DKVS), the vCenter Cluster state reconciles with the vSphere cluster when you restore your vCenter.

Cluster Information is saved in the ESXi hosts using DKVS and is synchronized with vCenter after a restore. This new feature will save administrators a lot of time.

And with this new feature in vCenter Enhanced Recovery of vCenter, we finish this first part of What’s New in vSphere 8.

In the second part, we will talk about:

• Unified Management for AI/ML Hardware Accelerators
• Virtual Hardware version 20
• Virtual TPM Provisioning Policy
• Migration aware applications
• High latency sensitivity with hyper-threading
• Simplified vNUMA configuration
• Compute Maximums
• Enhanced DRS Performance
• vSphere 8 security – Secure out of the box

Conclusion

We see in vSphere 8 many changes, not so significant, and the max resource limits did not change much (but may change until the official release). But overall, in this first part, we see a lot of good new features and improvements.

As I said at the beginning of this article, I think VMware is moving vSphere to a SaaS product. Some of the new features, the launch of vSphere+ and SAN+, are the beginnings of that path.

In this first part, I have highlighted the vSphere Distributed Services Engine, with the new DPU and GPU features.

Many improvements in the vSphere with Tanzu integration. With this fully Tanzu integration in vSphere 8, it will be easy to provide more workloads and implement and manage Tanzu Kubernetes Clusters.

The vSphere Lifecycle Manager also brings many changes, including how to stage and remediate ESXi hosts and the new DPU features.

And to finalize, one of my favorites in vCenter is, Enhanced Recovery of vCenter. To be able to back up vCenter, do some changes, and then when needed, restore vCenter without any issues and synchronize those after backup changes is something excellent and very useful.

Related Posts:

What’s New in vSphere 8 – Part 2
What’s New in vSphere 8 Update 1

Follow our Twitter and Facebook feeds for new releases, updates, insightful posts and more.