Windows Server 8 Hyper-V and High Availability

A few weeks ago we provided an overview of some of the new features and advancements we have made with Windows Server 8 Hyper-V. Today, I would like to turn the topic focus to High Availability (HA). As customers host more and more critical workloads in virtual environments, the need for integrated HA solutions for any size customer becomes increasingly important.

HA may mean different things to different types and size of organizations. In one example, if I am running a highly transactional workload, such as a storefront application, I may require 5x9 reliability. Many businesses run applications that are still critical to their needs, but they may not require this type of availability, or have the IT budget to support it. In our planning for Windows Server 8 Hyper-V we wanted to continue to deliver the best-in-class high availability solutions, and offer further options for the wide needs and budget of all types of businesses.

Windows Server Failover Clustering has been at the core of our solution offering, and we continue to build on this technology in Windows Server 8, adding many new features including:

Support for guest clustering via Fibre Channel. Windows Server 8 provides virtual Fibre Channel adapters within the virtual machine allowing your workloads access to storage area networks using fiber channel fabric. In addition, a virtual fibre channel enables you to:

  • Cluster guest operating systems over Fibre Channel providing HA for workloads within VMs
  • Use Windows multi-path I/O (MPIO) for high-availability and load balancing on the storage path

By employing MPIO and Failover Clustering together as complimentary technologies, users are able to mitigate the risk of a system outage at both the hardware and application levels.

Live migration enhancements. Live migrations have been enhanced to use more available network bandwidth which dramatically increases the performance of Live Migration and enables concurrent Live Migrations with no limits. The number of Live Migrations possible is governed by how much you want to invest in your network infrastructure.

Massive Scale. Windows Server 8 will now support up to 63 nodes and up to 4,000 virtual machines in a cluster.

Encrypted cluster volumes. Hyper-V, Failover Clustering and BitLocker now work in concert to create the ideal and secure platform for private cloud infrastructure. Windows Server 8 Cluster disks that are encrypted using BitLocker Drive Encryption enable better physical security for deployments outside secure data centers, providing a critical safeguard for the cloud and helping protect against inadvertent data leaks.

Cluster Shared Volume 2.0 (CSV). CSV has been greatly enhanced in a number of ways. From a usability standpoint:

  • CSV is now a core Failover Clustering feature (no longer a separate component that needs to be enabled)
  • Enabling CSV on a disk is now a single right click with the mouse
  • CSV disks are now included in the Failover Cluster Manager Storage view easing management

To support up to 63 nodes in a cluster, CSV has been improved in aspects of both performance and scalability. In terms of integrating with our partners, CSV has specifically been enhanced to work out of the box with storage filter drivers such as those used by: anti-virus, data protection, backup and storage replication.

Hyper-V application monitoring. With Windows Server 8, Hyper-V and Failover Clustering work together to bring higher availability to workloads that do not support clustering. It does this by providing a light-weight, simple solution to monitor application running in the VMs and integrating with the host. By monitoring services and event logs inside the virtual machine, Hyper-V and Failover Clustering can detect whether the key services that a virtual machine provides are healthy and provide automatic corrective action such as restarting the virtual machine or restarting a service within the VM.

Virtual machine failover prioritization. Virtual machine priorities can now be configured to control the order in which specific virtual machines failover or start. This ensures higher priority virtual machines are given the resources they need and lower priority virtual machines are given resources as they are available.

Inbox live migration queuing. Administrators can now perform large multi-select actions to queue live migrations of multiple virtual machines with ease and efficiency.

Affinity (and anti-affinity) virtual machine rules. Administrators can now configure partnered virtual machines so that at failover, the partnered machines are migrated simultaneously. For example, administrators can configure their SharePoint virtual machine and the partnered SQL Server virtual machine to always failover together to the same node. Administrators can also specify that two specific virtual machines cannot coexist on the same node in a failover scenario.

File Server Transparent failover. You can now more easily perform hardware or software maintenance of nodes in a File Server cluster (for example, storage virtual machine files such as configuration files, virtual hard disk files, and snapshots in file shares over the SMB2 protocol) by moving file shares between nodes with little interruption to server applications that are storing data on these file shares. Also, if a hardware or software failure occurs on a cluster node, SMB2 transparent failover lets file shares fail over to another cluster node with little interruption to server applications that are storing data on these file shares.

In our planning for expanded HA capabilities, we focused on two core goals:

  1. To provide the built-in ability to failover an entire data center
  2. While also providing affordable entry level HA options for customers.

To meet these goals, we developed a new functionality called Hyper-V Replica.

While Failover Clustering can be used to make a virtual machine highly available, it does not protect businesses from outage of an entire data center without the use hardware-based SAN replication across data centers. Hyper-V Replica fills an important customer need by providing an affordable in-box disaster recovery solution from an entire site, down to a single VM. It provides asynchronous, unlimited replication of your virtual machines over a network link from one Hyper-V host at a primary site to another Hyper-V host at a Replica site, without reliance on storage arrays or other software replication technologies.

clip_image002

In the event of failures at the primary site, the administrator can manually failover the production virtual machines to the Hyper-V server at the recovery site. During failover the virtual machines are brought online, and within minutes the rest of the network can access them successfully with minimal impact to the business. Once the primary site or virtual machines are restored, the administrators can easily revert the virtual machines back to the Hyper-V server at the primary site.

I have only touched on only a few of the core High Availability features that Windows Server 8 Hyper-V provides. You can see that with these enhancements we are building a virtualization platform that can support all your workloads and ensure the availability needs of your business.

David B. Cross

Director of Program Management

Windows Server Engineering