Many features are included in the Windows Server 2022 release for clustering. It is a rich feature set of high availability with greatly improved flexibility based on the needs of IT organizations. The new features relate to quorum behavior, virtual machine hosting, Active Directory–detached clusters, and a new dashboard.
Windows PowerShell Cmdlets for Failover Clusters As I have explained throughout this book, PowerShell is a command- line shell and scripting tool. Windows Server 2022 clustering has new cmdlets that provide powerful ways to script cluster configuration and management tasks. Windows PowerShell cmdlets have now replaced the Cluster.exe command- line interface.
Cluster Shared Volumes Cluster Shared Volumes (CSV) allows for the configuration of clustered virtual machines. CSV allows you to do the following:
■ Reduce the number of LUNs (disks) required for your virtual machines.
■ Make better use of disk space. Any VHD file on that LUN can use the free space on a CSV volume.
■ More easily track the paths to VHD files and other files used by virtual machines.
■ Use a few CSV volumes to create a configuration that supports many clustered virtual machines.
CSV volumes also are used for the Scale- Out- File-S erver cluster role.
Management of Large-S cale Clusters One advantage of Windows Server 2022 clusters is the ability for Server Manager to discover and manage the nodes in a cluster. By starting the Failover Cluster Manager from Server Manager, you can do remote multiserver management and role and feature installation. You can also manage a cluster from one convenient location.
Management and Mobility of Clustered Virtual Machines Microsoft, as well as the industry as a whole, is moving toward the cloud and virtualization. With that in mind, you can now configure settings such as prioritizing the starting or placement of virtual machines in the clustered workloads. This lets you allocate resources efficiently to your cluster.
Cluster- Aware Updating One issue that every administrator has dealt with is updating a system or application while it is running. For example, if you are running Microsoft Exchange and you want to do an Exchange update, when do you take the server offline to do the update? It always seems that someone is on the system 24 hours a day. Well, Windows Server 2022 clustering has a solution. Cluster-A ware Updating (CAU) is an automated feature that allows system updates to be applied automatically while the cluster remains available during the entire update process.
Cluster Node Fairness The Virtual Machine Load Balancing feature is new to Windows Server 2022. This new load-b alancing feature helps optimize the nodes in a cluster. When an organization builds a virtual machine cluster, there will be times when that cluster needs to have maintenance and certain virtual machines will be taken offline. When this happens, an unbalanced cluster (this is when some nodes are hosting VMs more often than others) may occur. This is where the VM Load Balancing feature (Node Fairness) helps the cluster. The Balancer will redistribute VMs from an over-b alanced node to an under- balanced node. To set up Node Fairness, use the PowerShell command (Get- Cluster).AutoBalancerLevel = <value>. The value input is 1, 2, or 3. 1 is equivalent to the Low setting (move the host when showing more than 80 percent loaded), 2 is equivalent to Medium (move the host when more than 70 percent loaded) and 3 is equivalent to High (average nodes and move the host when showing more than 5 percent above the average).
Cluster Operating System Rolling Upgrade One of the problems that many IT people face is the issue with downtime while their servers get upgraded to a new operating system. Windows Server 2022 includes a new feature called Cluster Operating System Rolling Upgrade. This new feature lets you upgrade a Hyper-V or Scale- Out File Server cluster from Windows Server 2012 R2 to Windows Server 2022 without stopping the servers.
Scale- Out File Server for Application Data By using Microsoft Storage Spaces, you can create a highly available clustered file share that uses SMB 3.0 and CSV to provide scalable access to data.
Scale- out file servers are useful for storing the following application data:
■ Hyper-V virtual machine storage
■ SQL Server database files
Be aware that scale- out file servers are not useful at all for typical file share data because they benefit only from applications that require a persistent connection to their storage. Shared Virtual Hard Disks In the previous versions of Windows, Failover Cluster nodes running as virtual machines had to use iSCSI or virtual HBAs to connect directly to SAN- based storage. With Windows Server 2022, you can set your Hyper-V virtualized cluster to use a shared VHDX virtual disk. Shared virtual hard disks can reside on the following:
■ A scale- out file server failover cluster
■ Cluster CSV volumes
Shared virtual hard disks are extremely useful in providing highly available shared storage for the following virtualized workloads:
■ SQL Server
■ Virtual Machine Manager
■ Exchange Server
Virtual Machine Drain on Shutdown When needing to perform maintenance on a Hyper- V failover cluster, you may have a lot of virtual machines on one node of a cluster. Inevitably, you will need to restart a cluster node for updates or shut it down for maintenance.
In previous versions of Windows, virtual machines running on the cluster would save their state, and then the cluster node would shut down. Windows Server 2022 helps alleviate this issue by automatically draining the virtual machines running on a node before it shuts down or restarts. Windows does this by attempting to live- migrate all virtual machines on the cluster node to other nodes in the cluster when at all possible.
This feature is turned on by default, but you can disable it using PowerShell.
Active Directory–Detached Clusters Previous versions of Windows Failover Clustering have depended on Active Directory to provide computer objects for the cluster name object as well as virtual computer objects. With Active Directory–detached failover clusters, communication to the cluster- form clients will use NTLM authentication rather than the normal Kerberos authentication. This is useful in maintaining high availability should a person accidentally delete a virtual computer object in Active Directory that a clustered resource depends on for Kerberos authentication.
Dynamic Witness Earlier in this chapter, I mentioned the dynamic quorum model and how votes were dynamically adjusted based on the number of nodes in a cluster. In Windows Server 2022, there is a new feature called dynamic witness that is enabled by default when the cluster is configured to use a dynamic quorum. Since it is preferred to have an odd number of votes at any one time in a cluster, the dynamic witness will turn on or off the witness vote in order to ensure that there are an odd number of votes in the cluster.
Tie Breaker For 50% Node Split Like the dynamic witness feature just described, the Tie Breaker For 50% Node Split option in Windows Server 2022 dynamically adjusts cluster node votes in order to maintain an odd number of votes in a cluster where no witness is being used. This is useful for a cluster in a site-a ware, stretched, or geocluster configuration.
Global Update Manager Mode Since the first release of Microsoft Cluster Services appearing in Windows NT 4.0 Enterprise, all nodes in a cluster maintain a local database that keeps a copy of the cluster configuration. The Global Update Manager (GUM) is a component of the cluster that ensures that before a change is marked as being committed for the entire cluster, all nodes have received and committed that change to their local cluster database. If one or more nodes do not report back or commit a change, the cluster node is kicked out of being a member of the cluster. Another issue that can occur is that for various clustered applications, such as SQL and Exchange, their performance can be negatively impacted by the time it takes the GUM to coordinate with all the nodes of a cluster for any changes. The GUM is only as fast as the slowest node in the cluster.
With Windows Server 2022, a new feature was added to Failover Clustering called Global Update Manager mode. This feature allows you to configure the GUM read- write modes manually in order to greatly speed up the processing of changes by the GUM and to improve the performance of certain clustered resources.
Turn Off IPsec Encryption For Inter- node Cluster Communications In network environments where IPsec is used, slow Group Policy updates and other issues can cause Active Directory Domain Services to be temporarily unavailable to cluster nodes. If the cluster intracluster communications protocol uses IPsec encryption, then these delays could cause cluster nodes to drop out of the cluster for failure to communicate in a timely manner with the rest of the nodes in the cluster. Windows Server 2022 provides a way to turn off IPsec encryption on the cluster communication network.
Cluster Dashboard Starting with Windows Server 2012, Failover Clustering supports up to 64 nodes in a cluster. Keeping track of the status and resources on all of these nodes can be an administrative headache! Managing more than one failover cluster and determining what a certain cluster hosts can be painful as well. Fortunately, in Windows Server 2022, the Failover Cluster Manager’s main dashboard has been updated to make it easier to see the status and health of multiple clusters.
Hyper- V Replica Broker Starting with Windows Server 2012, Hyper- V supported continuous replication of virtual machines to another server or cluster for disaster recovery purposes. The Hyper- V Recovery Broker allows virtual machines in a cluster to be replicated. The Hyper-V Recovery Broker keeps track of which cluster nodes virtual machines are residing on and ensures that replication is maintained.
Hyper- V Manager Integration into Failover Cluster Manager In Windows Server 2022, the Hyper- V Management Console is integrated with Failover Cluster Manager for managing virtual machines that are clustered. Normal Hyper-V operations such as configuring, exporting, importing, configuring replication, stopping, starting, and live- migrating virtual machines are supported directly through Failover Cluster Manager.
Virtual Machine Monitoring Starting with Windows Server 2012, Failover Clustering supports Virtual Machine Monitoring for Windows Server virtual machines. Virtual Machine Monitoring monitors administrator- selected Windows services running within a virtual machine and will automatically restart a service if it should fail. If the service does not start for the configured number of restart attempts, the virtual machine will fail over to another node and then restart. For example, you can configure Failover Clustering to monitor the Print Spooler service on a Windows Server 2022 virtual machine. If the Print Spooler service goes offline, then the cluster will attempt to restart the Print Spooler service within the virtual machine. If the service still fails, Failover Clustering will move the virtual machine to another node.
