Website under construction

Cluster operating system rolling upgrade 
Windows Server 2016 introduces a wonderful new method for upgrading the operating system of 
your server clusters with no downtime and dramatically reduced effort. This feature is called the 
cluster operating system rolling upgrade. Cluster operating system rolling upgrades initially will be 
limited to Hyper-V clusters and scale-out file servers (SOFS) clusters for Windows Server 2016. 
Until now, the cluster administrator was tasked with developing a detailed migration plan to update 
clusters with a new operating system. Often, administrators waited to move a cluster until new 
hardware was brought in as part of a system refresh. This often meant several years without any new 
capabilities for the cluster as well as some planned downtime for moving services between the old 
and new cluster. 
Cluster operating system rolling upgrade does not require the purchase of any additional hardware; 
the upgrades are done in place to each of the nodes. The cluster itself never needs to be stopped or 
restarted; the work takes place at the cluster node level, and all services remain online during the 
rolling upgrade process. Unlike typical cluster migration strategies, you do not need to make a new 
cluster. The existing cluster objects, including cluster name and cluster IPs, remain the same and 
online during the upgrade. Even better news is that you can fully reverse the process until the Cluster 
Functional Level attribute is changed. (More on that later.) 
Figure 2-37 shows a three-node Hyper-V cluster named ContosoPVTCloud. 

---

40 
CHAPTER 2 | Software-defined datacenter 
Figure 2-37:
The Failover Cluster Manager console with a three-node cluster 
Figure 2-38 and Figure 2-39 show three cluster shared volumes and a couple of VMs running.