Applies To: Windows Server 2016
This topic explains the new and changed functionality in Failover Clustering in Windows Server 2016.
Cluster Operating System Rolling Upgrade
A new feature in Failover Clustering, Cluster Operating System Rolling Upgrade, enables an administrator to upgrade the operating system of the cluster nodes from Windows Server 2012 R2 to Windows Server 2016 without stopping the Hyper-V or the Scale-Out File Server workloads. Using this feature, the downtime penalties against Service Level Agreements (SLA) can be avoided.
What value does this change add?
Upgrading a Hyper-V or Scale-Out File Server cluster from Windows Server 2012 R2 to Windows Server 2016 no longer requires downtime. The cluster will continue to function at a Windows Server 2012 R2 level until all of the nodes in the cluster are running Windows Server 2016. The cluster functional level is upgraded to Windows Server 2016 by using the Windows PowerShell cmdlt
- After you update the cluster functional level, you cannot go back to a Windows Server 2012 R2 cluster functional level.
- Until the
Update-ClusterFunctionalLevelcmdlet is run, the process is reversible, and Windows Server 2012 R2 nodes can be added and Windows Server 2016 nodes can be removed.
What works differently?
A Hyper-V or Scale-Out File Server failover cluster can now easily be upgraded without any downtime or need to build a new cluster with nodes that are running the Windows Server 2016 operating system. Migrating clusters to Windows Server 2012 R2 involved taking the existing cluster offline and reinstalling the new operating system for each nodes, and then bringing the cluster back online. The old process was cumbersome and required downtime. However, in Windows Server 2016, the cluster does not need to go offline at any point.
The cluster operating systems for the upgrade in phases are as follows for each node in a cluster:
- The node is paused and drained of all virtual machines that are running on it.
- The virtual machines (or other cluster workload) are migrated to another node in the cluster.The virtual machines are migrated to another node in the cluster.
- The existing operating system is removed and a clean installation of the Windows Server 2016 operating system on the node is performed.
- The node running the Windows Server 2016 operating system is added back to the cluster.
- At this point, the cluster is said to be running in mixed mode, because the cluster nodes are running either Windows Server 2012 R2 or Windows Server 2016.
- The cluster functional level stays at Windows Server 2012 R2. At this functional level, new features in Windows Server 2016 that affect compatibility with previous versions of the operating system will be unavailable.
- Eventually, all nodes are upgraded to Windows Server 2016.
- Cluster functional level is then changed to Windows Server 2016 using the Windows PowerShell cmdlet
Update-ClusterFunctionalLevel. At this point, you can take advantage of the Windows Server 2016 features.
For more information, see Cluster Operating System Rolling Upgrade.
Storage Replica is a new feature that enables storage-agnostic, block-level, synchronous replication between servers or clusters for disaster recovery, as well as stretching of a failover cluster between sites. Synchronous replication enables mirroring of data in physical sites with crash-consistent volumes to ensure zero data loss at the file-system level. Asynchronous replication allows site extension beyond metropolitan ranges with the possibility of data loss.
What value does this change add?
Storage Replica enables you to do the following:
Provide a single vendor disaster recovery solution for planned and unplanned outages of mission critical workloads.
Use SMB3 transport with proven reliability, scalability, and performance.
Stretch Windows failover clusters to metropolitan distances.
Use Microsoft software end to end for storage and clustering, such as Hyper-V, Storage Replica, Storage Spaces, Cluster, Scale-Out File Server, SMB3, Data Deduplication, and ReFS/NTFS.
Help reduce cost and complexity as follows:
Is hardware agnostic, with no requirement for a specific storage configuration like DAS or SAN.
Allows commodity storage and networking technologies.
Features ease of graphical management for individual nodes and clusters through Failover Cluster Manager.
Includes comprehensive, large-scale scripting options through Windows PowerShell.
Help reduce downtime, and increase reliability and productivity intrinsic to Windows.
Provide supportability, performance metrics, and diagnostic capabilities.
For more information, see the Storage Replica in Windows Server 2016.
Cloud Witness is a new type of Failover Cluster quorum witness in Windows Server 2016 that leverages Microsoft Azure as the arbitration point. The Cloud Witness, like any other quorum witness, gets a vote and can participate in the quorum calculations. You can configure cloud witness as a quorum witness using the Configure a Cluster Quorum Wizard.
What value does this change add?
Using Cloud Witness as a Failover Cluster quorum witness provides the following advantages:
Leverages Microsoft Azure and eliminates the need for a third separate datacenter.
Uses the standard publicly available Microsoft Azure Blob Storage which eliminates the extra maintenance overhead of VMs hosted in a public cloud.
Same Microsoft Azure Storage Account can be used for multiple clusters (one blob file per cluster; cluster unique id used as blob file name).
Provides a very low on-going cost to the Storage Account (very small data written per blob file, blob file updated only once when cluster nodes' state changes).
For more information, see Deploy a Cloud Witness For a Failover Cluster.
What works differently?
This capability is new in Windows Server 2016.
Virtual Machine Resiliency
Compute Resiliency Windows Server 2016 includes increased virtual machines compute resiliency to help reduce intra-cluster communication issues in your compute cluster as follows:
Resiliency options available for virtual machines: You can now configure virtual machine resiliency options that define behavior of the virtual machines during transient failures:
Resiliency Level: Helps you define how the transient failures are handled.
Resiliency Period: Helps you define how long all the virtual machines are allowed to run isolated.
Quarantine of unhealthy nodes: Unhealthy nodes are quarantined and are no longer allowed to join the cluster. This prevents flapping nodes from negatively effecting other nodes and the overall cluster.
For more information virtual machine compute resiliency workflow and node quarantine settings that control how your node is placed in isolation or quarantine, see Virtual Machine Compute Resiliency in Windows Server 2016.
Storage Resiliency In Windows Server 2016, virtual machines are more resilient to transient storage failures. The improved virtual machine resiliency helps preserve tenant virtual machine session states in the event of a storage disruption. This is achieved by intelligent and quick virtual machine response to storage infrastructure issues.
When a virtual machine disconnects from its underlying storage, it pauses and waits for storage to recover. While paused, the virtual machine retains the context of applications that are running in it. When the virtual machine's connection to its storage is restored, the virtual machine returns to its running state. As a result, the tenant machine's session state is retained on recovery.
In Windows Server 2016, virtual machine storage resiliency is aware and optimized for guest clusters too.
Diagnostic Improvements in Failover Clustering
To help diagnose issues with failover clusters, Windows Server 2016 includes the following:
Several enhancements to cluster log files (such as Time Zone Information and DiagnosticVerbose log) that makes is easier to troubleshoot failover clustering issues. For more information, see Windows Server 2016 Failover Cluster Troubleshooting Enhancements - Cluster Log.
A new a dump type of Active memory dump, which filters out most memory pages allocated to virtual machines, and therefore makes the memory.dmp much smaller and easier to save or copy. For more information, see Windows Server 2016 Failover Cluster Troubleshooting Enhancements - Active Dump.
Site-aware Failover Clusters
Windows Server 2016 includes site- aware failover clusters that enable group nodes in stretched clusters based on their physical location (site). Cluster site-awareness enhances key operations during the cluster lifecycle such as failover behavior, placement policies, heartbeat between the nodes, and quorum behavior. For more information, see Site-aware Failover Clusters in Windows Server 2016.
Workgroup and Multi-domain clusters
In Windows Server 2012 R2 and previous versions, a cluster can only be created between member nodes joined to the same domain. Windows Server 2016 breaks down these barriers and introduces the ability to create a Failover Cluster without Active Directory dependencies. You can now create failover clusters in the following configurations:
Single-domain Clusters. Clusters with all nodes joined to the same domain.
Multi-domain Clusters. Clusters with nodes which are members of different domains.
Workgroup Clusters. Clusters with nodes which are member servers / workgroup (not domain joined).
For more information, see Workgroup and Multi-domain clusters in Windows Server 2016
Virtual Machine Load Balancing
Virtual machine Load Balancing is a new feature in Failover Clustering that facilitates the seamless load balancing of virtual machines across the nodes in a cluster. Over-committed nodes are identified based on virtual machine Memory and CPU utilization on the node. Virtual machines are then moved (live migrated) from an over-committed node to nodes with available bandwidth (if applicable). The aggressiveness of the balancing can be tuned to ensure optimal cluster performance and utilization. Load Balancing is enabled by default in Windows Sever 2016 Technical Preview. However, Load Balancing is disabled when SCVMM Dynamic Optimization is enabled.
Virtual Machine Start Order
Virtual machine Start Order is a new feature in Failover Clustering that introduces start order orchestration for Virtual machines (and all groups) in a cluster. Virtual machines can now be grouped into tiers, and start order dependencies can be created between different tiers. This ensures that the most important virtual machines (such as Domain Controllers or Utility virtual machines) are started first. Virtual machines are not started until the virtual machines that they have a dependency on are also started.
Simplified SMB Multichannel and Multi-NIC Cluster Networks
Failover Cluster networks are no longer limited to a single NIC per subnet / network. With Simplified SMB Multichannel and Multi-NIC Cluster Networks, network configuration is automatic and every NIC on the subnet can be used for cluster and workload traffic. This enhancement allows customers to maximize network throughput for Hyper-V, SQL Server Failover Cluster Instance, and other SMB workloads.
For more information, see Simplified SMB Multichannel and Multi-NIC Cluster Networks.