What you will learn: VMware is the dominant hypervisor in most data centers today, but alternatives include Hyper-V from Microsoft, XenServer from Citrix and several Linux-based options. There is value in becoming a multi-hypervisor data center, as moving between hypervisors is now easier than ever. Analyst George Crump discusses those values and how to create a multi-hypervisor data center.
There are several reasons to use multiple hypervisors, but the primary one is simply cost savings. VMware's competitors, including Microsoft's Hyper-V, represent a significant cost savings versus the standard licensing fees from VMware. In the past, moving to these alternative hypervisors meant giving up a significant set of advanced features, such as live server migration and automated resource balancing. Now, while still not as feature-rich as VMware, these hypervisors have closed the gap and in some cases, have even added features that VMware does not have.
Some of these alternative hypervisors are also easier to use. For example, a Windows-only data center may find Hyper-V significantly easier to work with since it is integrated into the operating system. The same holds true for 100% Linux environments, such as those offered by cloud providers. For them, selecting a Linux-based hypervisor may be much simpler than integrating VMware.
Leveraging the cloud to host some of your computing needs may also be something you want to consider. Many large public cloud providers are not based on VMware. As a result, you may find yourself running on another hypervisor whether you want to or not.
What the multi-hypervisor advantage comes down to is flexibility. Most data centers should and will have VMware as their core hypervisor for mission-critical workloads. But being able to leverage Hyper-V or one of the Linux options may make sense to reduce costs or to take advantage of a specific feature that one hypervisor may have over the other. The key is to have the flexibility to move between these hypervisors as seamlessly as possible.
While manual movement of virtual machines (VMs) between multiple hypervisors is possible, it is not very flexible. As a result, most data centers should use some automated method of moving VMs.
Backup and recovery between multiple hypervisors
The most basic way to move between multiple hypervisors is to leverage a backup and recovery tool. This is especially easy when your backup application supports each hypervisor natively, such as Veeam or Dell AppAssure, for example. When it is time to move to a different hypervisor, you simply recover the VM to that hypervisor.
The challenge with this method is that it requires downtime for the VM while a final backup is executed and the recovery takes place. The time required to perform these two steps makes this option only viable for smaller VMs that can be moved during a planned maintenance window.
Replication between hypervisors
For situations where you want the movement between hypervisors to occur in real-time, replication between hypervisors is an ideal solution. Products such as Vision Solutions' Double-Take can run on multiple operating systems and hypervisors, as well as run as instances on cloud service provider platforms like Amazon EC2.
Armed with this type of tool, data centers can leverage alternate hypervisors for more than just planned move-and-maintenance situations. They can be part of a stand-by or bursting strategy. Since data is replicated in real-time, shutting down the VM on the primary hypervisor and then starting the VM on the secondary hypervisor is all that needs to happen.
An interesting use case for this type of capability is to leverage the alternate hypervisor for failover or disaster recovery instead of paying for two times the VMware license, then replicating to the less expensive hypervisor and using it as the "backup" cluster for all your critical VMs. Being able to replicate to the cloud takes this a step further by even eliminating the need to have stand-by hardware in the primary data center.
Integrating hypervisors into VMware
Applications like HotLink Software Hybrid Express integrate hypervisors directly into the vCenter console. This enables the movement of VMs between the supported platforms from within the primary hypervisor interface the administrator is accustomed to using. This also allows administrators to drag and drop VMs from one hypervisor or cloud to another. Access to all the features are triggered from the same VMware-type of mouse commands. The net impact is seamless migration and management of VMs between platforms without having to learn a new interface.
The concept of a commoditized hypervisor is not new. Free hypervisors have been available for years, including from VMware. Leveraging lower-cost hypervisors, especially now that they are closing the feature gap, is a critical capability for all data centers. What makes this multi-hypervisor data center a reality is the ability to quickly and transparently move data between the hypervisor types.
Working with VMware's vCenter Manager
Keys to multi-hypervisor management
Overcoming challenges of multi-hypervisor environments
Dig Deeper on Virtualization Strategy
George Crump asks:
Do you think the benefits of a multi-hypervisor environment outweigh the challenges?
0 ResponsesJoin the Discussion