Newer versions of software usually bring a combination of new features, bug fixes, and “under-the-hood” improvements. VMware vSAN 6.7 is no exception. Hyperconverged Infrastructure (HCI) environments running older versions of vSAN such as 6.0, 6.2, and 6.5 will benefit from these enhancements after upgrading to the latest version. This article covers 10 reasons to get started on those upgrades.
It is important to note that vCenter Server should be upgraded to the latest version before upgrading the hosts (vSphere and vSAN). As with any software upgrade, you should ALWAYS read the release notes before performing the upgrade and follow the recommendations in the documentation. Upgrading vCenter Server from an older version to the latest brings updates to VMware vSphere Update Manager, which is our first reason for upgrading.
vSphere Update Manager
vSphere Update Manager (VUM) functionality is incorporated into the modern HTML5 “Clarity” UI that is used to manage vSphere and vSAN. This provides tighter integration with the rest of the vSphere UI, which simplifies operational activities such as patching and upgrades.
VUM includes a built-in “vSAN Cluster” baseline, which combines information from the hardware in use, the VMware Release Catalog, and the vSAN hardware compatibility list (HCL) to provide guidance on what vSphere and vSAN versions should be running in the cluster. This takes the guesswork out of understanding what updates, if any, should be done to provide the best experience. VUM also includes the ability to update firmware and drivers for certain storage controller makes and models used with vSAN. A central location to update vSphere, vSAN, firmware, and drivers increases efficiency and minimizes risk.
Guided cluster creation and extension
Organizations deploying new HCI clusters with vSAN will benefit from Cluster QuickStart. This new feature streamlines the process of setting up a vSAN cluster. Several configuration items are checked after hosts are added to the cluster including software version compatibility, compliance with the vSAN HCL, certified controller driver and firmware versions are in place, and time is synchronized across the environment.
Items that do not pass the check are flagged so that you can correct the issues before configuring the cluster. Validating the environment after issues are resolved is done with a single click.
Configuring the cluster is also much easier with Cluster QuickStart. Previous versions of vSphere and vSAN required a number of manual steps to set up a cluster. It is now possible to configure items such as virtual distributed switches, vMotion, and vSAN port groups, network configurations, vSphere HA and DRS, and vSAN disk groups in a matter of minutes. Cluster QuickStart also supports more advanced vSAN setup options such as stretched clusters, encryption, and fault domains. Expanding a cluster through the addition of hosts is also faster and easier using Cluster QuickStart. A click-through demonstration of Cluster QuickStart is located here.
Perhaps the most important benefit of Cluster QuickStart is consistency. An infrastructure that is configured the same according to vendor best practices helps ensure the highest levels of availability and performance. Cluster QuickStart simplifies and standardizes the configuration of an HCI cluster powered by vSAN to provide the best possible experience.
vRealize Operations enhancements within vCenter Server
The latest versions of vRealize Operations feature built-in dashboards specifically for vSAN environments. These dashboards provide detailed views for a variety of tasks including capacity monitoring, operations, and troubleshooting. Simplified dashboards are also included directly in the vSphere Client. This makes it easy for you to see key information such as alerts, capacity information, and performance data at a glance without leaving the vSphere Client.
A more complete description of this functionality is found in this blog article: vSAN 6.7 and vRealize Operations within vCenter.
De-staging and resynchronization improvements
Some of the best enhancements are “under the hood.” VMware engineers that work on vSAN continue to make enhancements to vSAN performance with each release. In some cases, these changes are not apparent. In other cases, they can make significant, positive impacts on workloads running on vSAN.
The latest versions of vSAN include optimization of de-staging data from the cache tier to the capacity tier of a vSAN datastore. The speeds at which these de-staging operations occur has increased resulting in improved data resynchronization and virtual machine IO performance—especially for write-intensive workloads.
Speaking of resynchronization, the method used to perform this activity is also improved in vSAN 6.7 and newer versions. This action is now performed using an adaptive resynchronization mechanism that helps reduce the amount of time to achieve storage policy compliance when a drive or host is offline. A minimum amount of bandwidth is guaranteed to resynchronization activities when there is resource contention. When there is free bandwidth, additional resources can be allocated to speed up the resynchronization process. This technote provides more details on vSAN’s adaptive resynchronization feature: Adaptive Resync.
Maintenance mode enhancements
vSphere hosts should be placed into maintenance mode when performing actions such as software upgrades, replacing faulty hardware, and adding drives. Compute resources and storage capacity are reduced when an HCI host enters maintenance mode. This is because vSAN uses local drives that are installed in the hosts. A reduction in compute and storage resources can introduce risk in clusters that are heavily utilized if not managed properly. Enhancements to the latest version of vSAN help mitigate this risk.
vSAN now performs a simulation of data evacuation from a host going into maintenance mode to determine there is sufficient capacity on the other nodes to house the evacuated data. vSAN predicts whether the operation will be successful. If not, vSAN halts the host’s entry into maintenance mode before data is moved and produces an alert. This helps prevent scenarios where vSAN could run out of free space—a bad thing for any storage solution.
Improved support experience
vSAN Support Insight uploads anonymous data from your vSAN environment to the VMware analytics cloud when the VMware Customer Experience Improvement Program (CEIP) feature is enabled. Configuration, feature usage, performance, and product log data are uploaded and accessible by VMware Global Support Services (GSS). A technical support engineer (TSE) has access to this anonymous data when you provide your vCenter Server UUID to GSS. This enables the TSE to begin the troubleshooting an issue right away without the need to manually upload log bundles. TSE access to this data can speed up problem resolution times considerably. The image below shows an example of what the TSE can see. Notice the hostnames, storage policy names, and so on are random values.
Another benefit to enabling CEIP is access to vSAN online health checks. These health checks are pushed down to vSAN environments when new issues, recommendations, and so on are found. vSAN Health is dynamically updated to include new checks without the need to install additional software. These checks provide guidance and warnings to help you configure and operate vSAN according to VMware best practices.
VMware can also use this data to better understand how customers use VMware solutions such as vSAN. Information on product usage and trends helps VMware make continuous improvements and discover potential issues before they become a problem.
Learn more about vSAN Support Insight in this technote.
Historical and usable capacity reporting
This topic might be less exciting than previous items we have covered, but it is important. Monitoring and managing storage capacity in any environment are critical parts of daily operations. vSAN continues to make small but notable improvements when it comes to capacity reporting.
vSAN 6.7 Update 1 introduced a historical capacity dashboard that reports usage over a period of time. This includes changes to the deduplication and compression ratio. Historical data enables you to identify trends and more accurately predict when additional capacity might be needed. This enables IT to be more agile and responsive to business needs.
This release also includes a usable capacity (as opposed to raw capacity) estimator. You can select a storage policy and vSAN will display the estimated amount of usable capacity based on that policy. The capacity consumption difference between mirroring and erasure coding storage policies, as an example, is easy to see using this feature.
vSphere and vSAN FIPS 140-2 validation
Security is top-of-mind for nearly all organizations. vSAN 6.7 encryption has been validated for the Federal Information Processing Standard (FIPS) 140-2. FIPS validated software modules have numerous advantages over special purpose hardware because they can be executed on a general-purpose computing system, providing portability and flexibility. You can configure a vSAN host using any HCL-compatible set of drives in thousands of form factors, capacities and features, while maintaining data security using FIPS 140-2 validated modules. This VMware Security web page provides more details on FIPS validation.
TRIM/UNMAP for storage space efficiency
The latest version of vSAN is capable of automated space reclamation with TRIM and SCSI UNMAP support. SCSI UNMAP and the ATA TRIM command enable a guest OS file system to notify a storage platform that a block is no longer in use. Blocks can be reclaimed to free up storage capacity. This automated process helps you get the most capacity out of your vSAN clusters resulting in a lower total cost of ownership. This blog article provides more information.
vSAN Stretched cluster improvements
Stretched clusters with vSAN minimize the cost and complexity of providing high availability across sites. The option to deploy a vSAN stretched cluster configurations has been around for a few years. vSAN 6.7 and newer versions contain improvements specifically for stretched clusters.
One of the updates is more efficient inter-site resynchronization. vSAN includes the ability to make data redundant across sites and locally within each site. vSAN 6.7 and newer versions send only one copy of the data across sites when a resync is required. Redundancy within a site is then created from that single copy. This approach minimizes the amount of inter-site bandwidth consumed by resync traffic.
vSAN Witness traffic separation provides the option to configure a dedicated VMkernel NIC for witness traffic. This option improves security as vSAN data is not transmitted across that VMkernel NIC. It also provides better flexibility for stretched cluster network configuration as witness traffic has less stringent bandwidth and latency requirements than vSAN traffic between data sites.
Site continuity is also more intelligent. vSAN 6.7 and newer versions will determine which site will produce the highest level of availability when there is a network partition between the preferred and secondary sites. The secondary site can operate as the active site until the preferred site has the latest copy of the data. This prevents the VMs from migrating back to the preferred site and losing locality of data reads.
The vSAN Stretched Cluster Guide is a great resource for those deploying and managing vSAN stretched clusters.
It is well understood that most software upgrades introduce a fair amount of planning, effort, and a bit of risk. However, the benefits of upgrading to the latest version of vSAN are clear. The 10 items listed in this article are significant enhancements. There are also numerous other improvements, bug fixes, performance tweaks, and so on—too many to list in a blog article like this. The release notes for each version contain these details. As mentioned previously, it is best to review those release notes and the documentation thoroughly before performing an upgrade.