VMware is gearing up for a significant update with vSphere 8 Update 2, and it’s set to make waves in the realm of virtualization. Anticipated for release in Q3 2023, this update promises to bring exciting changes that will have a positive impact on the daily routines of VMware administrators.
Reduced downtime upgrades
One of the most striking and game-changing features in this update is the remarkable reduction in downtime during upgrades. Initially introduced as part of the vSphere Plus offering, this feature is now extending its reach to on-premises instances that may not be cloud-connected. This expansion of functionality is a significant boon for administrators, as it empowers them to update or patch their vCenter with just a few minutes of downtime.
In the fast-paced world of virtualization, staying up-to-date is paramount. However, the challenge lies in coordinating downtime, especially when there are over seven patches to implement annually. In medium-sized environments, each upgrade can consume a precious hour of downtime. It’s essential to note that this improvement doesn’t currently apply to vCenters in ELM (Enhanced Linked Mode) and HA (High Availability) configurations. But the good news is that now, with the latest update, vCenter upgrades promise to be a breeze, with downtime reduced to a mere five minutes.
This remarkable feat is achieved through a well-thought-out process. When it’s time to upgrade, a new vCenter server is deployed with a temporary IP address, a practice that holds true even for minor patches. Subsequently, data is seamlessly copied from the old vCenter to the new one. The only brief downtime occurs during the switchover process and the subsequent startup of services. Adding to the resilience of this approach, an automatic LVM (Logical Volume Manager) snapshot is taken during the patching process. This snapshot can be conveniently resumed if a failure occurs, or it can be rolled back if necessary. This capability to resume an upgrade is especially valuable because sometimes an upgrade fails due to a minor hiccup, and having the ability to fix the issue and resume the upgrade process is a significant time-saver.
Certificate management has long been a less-than-joyful task, but over the past few years, there have been strides in making it more user-friendly. In the latest update, there’s a particularly notable enhancement that’s poised to make everyone’s lives easier: vCenter certificates can now be renewed or replaced without any need for service restarts, ensuring uninterrupted operations.
This improvement is particularly timely given the growing emphasis on certificate management, with web browsers increasingly shortening the validity period of certificates they accept. Recognizing this challenge, vSphere has incorporated a highly beneficial feature. Now, administrators can seamlessly renew or replace vCenter certificates without the inconvenience and potential downtime associated with restarting services.
This enhancement not only streamlines the certificate management process but also ensures that vSphere remains in step with evolving security and compliance requirements. It’s a testament to vSphere’s commitment to delivering a more efficient and hassle-free experience for its users
Ensuring the reliability of network recovery is paramount, especially in complex virtualized environments. In this latest update, vSphere introduces a significant enhancement in this area, simplifying the restoration process for environments that utilize one or more vSphere distributed switches (vDS).
Traditionally, when restoring vCenter from an outdated backup, it could be a challenge to ensure that it seamlessly reconciled with the current version of the distributed switch on ESXi hosts. This could lead to potential inconsistencies and complications. However, vSphere’s latest improvement addresses this issue comprehensively.
This enhancement becomes particularly valuable in scenarios where you find yourself in a precarious situation, with your vDS not synchronized across all hosts. Such situations often arise when a backup is taken, followed by changes to the vDS configuration, and then a subsequent restore is necessary. The good news is that with this update, those concerns about vDS inconsistencies when restoring from a backup are a thing of the past. Now, vDS changes are intelligently and automatically propagated from clusters to the vCenter, ensuring seamless synchronization. What’s more, this capability extends to environments using vDS integrated with NSX, further enhancing its versatility and utility.
Enhanced Security with vSphere Identity Federation
Undoubtedly, security remains at the forefront of vSphere’s priorities, and it continues to innovate its identity federation features with this objective in mind. Since the advent of vSphere 7, there’s been a steady augmentation of supported identity providers. With the transition to vSphere 8, the inclusion of support for OKTA identity services marked a significant stride forward. However, it’s the latest Update 2 release that brings a pivotal transformation, coinciding with Microsoft’s recent rebranding of Azure AD to Entra ID.
This rebranding aligns seamlessly with vSphere’s dedication to adaptability. Now, Entra ID takes its place among the supported identity providers, alongside the existing array, ensuring a wide spectrum of choices for multi-factor authentication and robust security layers.
But that’s not all – there’s a noteworthy enhancement on the Microsoft side. In the latest update, a feature is introduced that simplifies Active Directory Organizational Unit (OU) path management during VM customization. Now, within the VM customization wizard, you have the option to add an AD OU path directly. This eliminates the need for computer objects to default to the “Computers” container or require external automation. It’s a streamlined solution that underscores vSphere’s commitment to user-friendly and efficient identity management.
Enhanced vSphere Lifecycle Manager (vLCM)
The vSphere Lifecycle Manager (vLCM) has been a transformative force, revolutionizing how many manage their virtual environments. In its continual pursuit of improvement, vSphere 8 Update 2 introduces significant enhancements to vLCM’s already impressive capabilities. While it has historically provided support for vSAN witness nodes and vSAN clusters, this latest update ushers in a noteworthy change. It grants vLCM the capability to oversee and manage witness nodes participating in multiple vSAN clusters.
This update represents a pivotal shift in the management of shared vSAN witness nodes. Users now have the ability to independently manage the image definition of these witness nodes, regardless of the vSAN clusters they are associated with. This newfound flexibility is a game-changer, particularly for scenarios involving shared vSAN witness nodes. It empowers administrators to tailor image builds to suit specific use cases, ensuring that the virtual environment is precisely configured to meet the unique requirements of each cluster.
The vSphere Configuration Profiles feature, initially introduced in vSphere 8 and subsequently refined in vSphere 8 Update 1, continues to evolve and improve in Update 2. This latest iteration introduces a user-friendly and comprehensive UI workflow that simplifies the creation, editing, and application of vSphere Configuration Profiles.
One notable enhancement is the elimination of the need to export the JSON document for editing purposes, although this option is still available for advanced users. Instead, a new ‘Draft’ tab graces the user interface, offering an intuitive platform for creating, editing and applying drafts or copies of existing configurations. For instance, administrators can effortlessly augment NTP (Network Time Protocol) settings within a cluster configuration document. The improved interface streamlines the process, allowing edits to be saved in a draft format and applied in a manner akin to host profiles today.
Windows VM Deployment with Streamlined Processes
While it may appear to be a minor adjustment, a significant improvement has been implemented in the deployment of Windows VMs. Users can now specify the Organizational Unit (OU) path during the creation of customization specifications. This seemingly subtle change carries substantial benefits, ensuring that Windows VMs are not just deployed but also customized to align precisely with the specified OU path. This refinement greatly enhances the efficiency and precision of VM deployment and integration into Active Directory.
Descriptive Error Messages
Significant improvements have been made to error messages, addressing a long-standing source of frustration for users. This enhancement aims to provide greater clarity and utility when troubleshooting issues. One notable improvement is the transformation of error messages related to locked VM files.
In scenarios where a VM cannot be powered on, the updated error messages will now precisely pinpoint the locked file and identify the host responsible for the lock. This change is a significant leap forward in resolving issues efficiently and effectively. We’ve all experienced the challenges of tracking down file locks on VMDKs, often resorting to laborious CLI commands and sifting through extensive logs. But those days of frustration are now behind us. With these detailed error messages, you’ll receive specific information, including the IP address and MAC of the host holding the file lock. This newfound clarity is a valuable asset in swiftly diagnosing and resolving file lock issues, simplifying the troubleshooting process, and minimizing downtime.
Expanded DPU Ecosystem
In the vSphere 8 release, VMware introduced support for Data Processing Units (DPUs) within its distributed services engine, fostering collaborative efforts with industry giants such as Dell, HP, Nvidia, AMD, and Intel. This collaborative initiative aimed to bring enhanced capabilities to virtualized environments.
Excitingly, VMware is now taking a step further in its commitment to expansion and partnership. The support for DPUs is being extended to include servers from Lenovo and Fujitsu, underscoring VMware’s determination to continually broaden its partner ecosystem. This expansion signifies VMware’s dedication to providing a wider range of options and solutions, ultimately benefiting its user base and advancing virtualization technology.
AI And ML Getting Bigger And Bigger
The significance of GPUs within vSphere continues to grow, with a steadfast commitment to treating them as first-class citizens. Subsequent releases have witnessed substantial enhancements in GPU support, marking a pivotal shift in the virtualization landscape. For example, vSphere now boasts the capability to vMotion GPU-enabled virtual machines, providing more flexibility and agility.
In the latest vSphere 8 Update 2 release, an exciting addition takes center stage: smarter placement of GPU-enabled VMs, thanks to an upgraded Distributed Resource Scheduler (DRS). This revamped DRS is designed to make superior initial placement decisions and offers the option for automatic load balancing of vGPU-enabled virtual machines. As the realms of Artificial Intelligence (AI) and Machine Learning (ML) continue to expand, it becomes increasingly crucial to have robust GPU features at the hypervisor level. Previously, there were several limitations associated with VMs equipped with vGPUs. However, vSphere’s latest improvements in vGPU placement address these concerns comprehensively. DRS now excels in making optimal deployment decisions, especially when it comes to the initial placement of a VM. Furthermore, vGPU-enabled VMs are intelligently and automatically migrated when necessary to accommodate larger workloads. This capability ensures that virtual environments can efficiently adapt to evolving demands.
In addition to these enhancements, there’s one more noteworthy addition that enhances the vGPU experience. No longer will users need to guess how much downtime a vMotion operation will entail. The ability to view a stun time estimate directly in the VM’s settings provides valuable insight, helping administrators make informed decisions and minimize disruptions.
Improved DRS Featuring vGPU Optimization
In previous vSphere iterations, VMs with specific GPU demands often encountered placement challenges, particularly when the required GPUs were spread across different hosts. Enter vSphere 8 Update 2, featuring an advanced Distributed Resource Scheduler (DRS) capability designed to automatically optimize the placement of vGPU-enabled VMs.
This marks a significant step forward in bolstering GPU workload support. Now, in scenarios where an incoming VM necessitates, for instance, four GPUs and those resources aren’t available on the same host, DRS springs into action. It intelligently orchestrates the migration of another VM to create the required space. This dynamic load balancing ensures that VMs are promptly and appropriately placed and powered, eliminating resource bottlenecks and enhancing overall performance.
Quality of Service for GPU Workloads
In the realm of virtual GPUs (vGPUs), the “stun time,” which refers to the brief pause a VM experiences during migrations, can sometimes be a significant concern. However, with the introduction of vSphere 8 Update 2, administrators gain a valuable tool to assess and manage this aspect. Specifically, they can now access a detailed estimate of the maximum stun time for vGPU-enabled VMs, a calculation based on network speed and vGPU memory size.
This addition provides administrators with a deeper level of control and understanding. Moreover, it empowers them to define a “Quality of Service” (QoS) for VMs by setting a maximum acceptable stun time. VMs that fall within this established QoS threshold can be subjected to automated load balancing or manual migration processes. However, in cases where a VM surpasses the predefined stun time, such as exceeding 10 seconds, its migration is restricted. This restriction is in place to ensure that VM operations remain undisturbed, prioritizing seamless performance and user experience.
More Vroooom For Your VM
Hardware updates are always eagerly anticipated, and the introduction of Virtual Machine Hardware Version 21 doesn’t disappoint. This latest iteration brings substantial enhancements to VM device capacities, catering to the evolving needs of virtualized environments:
- Increased vGPU Capacity: VMs can now harness the power of up to 16 vGPUs, unlocking greater potential for graphics-intensive workloads.
- Expanded NVMe Disk Attachments: Hardware Version 21 empowers VMs by allowing the attachment of up to 256 NVMe disks, providing ample storage resources for demanding applications.
- Support for NVMe 1.3 Specification: Windows users and Windows Server environments with NVMe disks benefit from compatibility with the NVMe 1.3 specification, ensuring optimal performance and efficiency.
- Enhanced Compatibility: Hardware Version 21 aligns seamlessly with popular operating systems, including Red Hat 10, Oracle 10, Debian 13, and FreeBSD 15, broadening its utility across diverse environments.
NOTE: It’s essential to note that to fully leverage these capabilities, you’ll need both vSphere 8 Update 2 and Hardware Version 21. Also, keep in mind that Hardware Version 21 is specific to ESXi 8 Update 2 and subsequent releases. Therefore, an immediate hardware update may not be necessary unless you’re actively seeking to capitalize on these new features. This strategic approach ensures that your virtual infrastructure remains current while optimizing resource allocation and performance as needed.
The Tanzu Ecosystem
In vSphere 8 Update 2, the deployment of Supervisor Clusters, a feature that has been integral since vSphere 7, is now more straightforward than ever. Administrators benefit from an import/export option that allows them to effortlessly export configurations into a readable JSON document. This document can be conveniently reused for deploying additional clusters, streamlining the process, and exemplifying vSphere’s support for DevOps practices.
To further enhance efficiency for those seeking to replicate configurations, a valuable quick clone feature is at your disposal. This feature simplifies the process of duplicating configurations from one vSphere cluster to another, facilitating the expansion of resources and TKG (Tanzu Kubernetes Grid) instances. This capability aligns perfectly with vSphere’s commitment to providing tools that streamline operations and promote flexibility within containerized environments.
The latest release of VMware vSphere 8.0 Update 2 is indeed substantial, bringing with it a multitude of impressive new features and capabilities. This evolution of the vSphere stack introduces a wealth of new “cloud-centric” functionalities, empowering organizations to embrace greater agility and equipping them with the essential tools to facilitate a cloud-first approach in their operations and workflows.
For further insights and details, I recommend checking out the official post here: Announcing vSphere Q3 2023 Release
🔥Subscribe to the channel: https://bit.ly/3vY16CT🔥
🚨Read my blog: https://angrysysops.com/
🛒 VMware EMEA store: https://imp.i263671.net/c/3505578/814646/11461
🛒 VMware US store: https://imp.i263671.net/c/3505578/814642/11461
🛒 VMware APAC store: https://imp.i263671.net/c/3505578/814645/11461https://youtu.be/U2G6krsK62w