Rawpixel - Fotolia


Hyper-V 2016 features support shared storage alternatives

Storage improvements, like support for file-based and block-based storage, help enhance Hyper-V 2016 features, including guest clustering and failover.

Hyper-V 2016 ships with a lot of new features and improvements to existing features, including storage. One of...

the main reasons to provide support for various types of storage is to ensure all types of organizations can take advantage of all the features Hyper-V 2016 offers.

Support for alternative shared storage is important because storage area networks (SANs) are expensive. To maintain VM availability without a SAN, Microsoft introduced iSCSI Target Server and Storage Spaces in Windows Server 2012. Similarly, in Windows Server 2016, Microsoft developed a new mechanism to help organizations use locally attached storage and build a scalable storage infrastructure to provide storage for both physical and virtual deployments.

File-based and block-based storages

Hyper-V 2016 provides support for both file-based storage and block-based storage. File-based storage includes file shares that have been created on Windows file servers or network-attached storage devices. Block-based storage includes storages that support both Storage Management Initiative Specification (SMI-S) and symmetric multiprocessing (SMP) storage specifications. Microsoft developed Windows Storage Management API, also known as SMAPI, in Windows Server 2012, which can be used to manage both SMI-S and SMP block-based storages.

Storage Spaces

In Windows Server 2012 R2, Microsoft introduced Storage Spaces, which is a good option if you need a flexible storage infrastructure. Storage Spaces provides a cost-effective, highly available and scalable storage infrastructure to provide storage for both virtual and physical deployments. You can group a bunch of hard disks together using Storage Spaces and then make those disks available for virtual and physical environments. Storage Spaces technology is closely integrated with Windows failover clustering. You can use Storage Spaces for Clustered Shared Volumes and Scale-Out File Server deployments. Storage Spaces supports continuous availability, mirroring, storage tiers and write-back cache. Although most organizations still use a SAN, many new deployments use virtual storage, which includes Storage Spaces as the back-end storage system, because of the flexibility it provides.

Storage Spaces Direct

In Windows Server 2016, Microsoft introduced a new storage technology called Storage Spaces Direct, which allows organizations to use locally attached Hyper-V storage to achieve workload availability. You can have up to 16 servers and more than 400 connected drives, providing 1,000 TB of storage per Hyper-V cluster. Storage Spaces Direct architecture is very flexible and scalable. If you need to scale out, all you need to do is add new drives.

In Windows Server 2016, Microsoft introduced a new storage technology called Storage Spaces Direct, which allows organizations to use locally attached Hyper-V storage to achieve workload availability.

To add a new drive, attach it to the Hyper-V host, and then let Windows OS discover it. Once discovered, Storage Spaces Direct will check to see if the drive can be added to storage pools by checking the CanPool property of the drive. If the CanPool property says True, the new drive will automatically be added to storage pools maintained by Storage Spaces Direct. By running the Get-PhysicalDisk | CanPool command, you can check the CanPool property of all the drives. Once drives have been discovered and added to the storage pools, you're ready to create new volumes. If the drive doesn't appear in storage pools, rescan drives using the Device Manager or format the drives, if need be.

Note that Storage Spaces Direct is an enhancement to Storage Spaces introduced in Windows Server 2012. The underlying architecture of Storage Spaces Direct implements a new file system called Resilient File System (ReFS), which supports various storage operations, like expansion, checkpoint merging and checksums, to detect and correct any drive errors. ReFS is also capable of rotating the data to hot or cold storage tiers based on data usage.

iSCSI Target Server

Although Microsoft introduced iSCSI Target Server in Windows Server 2012 to provide shared storage functionality for both Hyper-V hosts and VMs, the company doesn't recommend its use in production because it has high latency and might not be suitable for workloads that require low-latency storage systems. If you use iSCSI Target Server on a Windows Server 2012 or later OS, you can implement Hyper-V failover clustering for testing and development purposes. Simply use iSCSI clients on Hyper-V hosts or VMs to connect to the targets defined in the iSCSI Target Server.

Hyper-V 2016 features and storage

Some of the storage enhancements Microsoft introduced ensure that administrators can easily use Hyper-V 2016 features. For example, Microsoft provided guest clustering support that can be implemented without requiring you to expose shared storage to VMs. By using a simple VHDX file, you can set up clustering between VMs. Though guest clustering does require Windows Server 2012 R2 or later OSes, at least there's no need to expose shared storage from a SAN or equivalent storage device.

Next Steps

Understand different types of Hyper-V virtual switches

Use Hyper-V replication or Storage Replica

Manage Hyper-V VM performance with this feature

Dig Deeper on Microsoft Hyper-V management