Virtualization creates virtual resources that can be used the same way as any physical resource or application. Disk drive partitioning is a familiar form of virtualization. Servers, networks, desktops, applications, and storage can all be virtual.
Virtualization started at IBM in the 1960s as a way of partitioning large mainframe computers into separate virtual machines that could run various processes in a more flexible way. Even as mainframes gave way to desktops, servers, and workstations, virtualization evolved to help manage a variety of applications throughout a network of desktops and servers.
Virtualization is now an integral part of both desktop infrastructures and data centers. It has a variety of advantages, from flexibility to cost, and is easy and relatively cheap to implement. Data centers now represent a key business capability for executing many business initiatives, and improving their effectiveness shows good returns.
The evolution of virtualization continues. Not too long ago, one application per server was pretty much the rule. Now many virtual machines, with full workloads, can fit onto a single piece of hardware.
The ability to monitor and manage a large number of virtual machines has not kept pace with their proliferation. The need to manage the hardware and network does not go away, so managing virtual machines is an added responsibility. But without proper management, you might not realize the full benefits of virtualization, and, in fact, find yourself facing serious performance bottlenecks.
The Many Benefits of Virtualization
A virtual infrastructure can make enterprise-level technology affordable to organizations that lack the capital for hardware and software licenses or the cash flow for continuous data center maintenance. This approach to maximizing resource use is scalable, flexible, and allows for easier load balancing on servers.
Sharing hardware via virtualization reduces capital budgets. Physical hardware often sits idle for significant periods of time. When that same device stands in for multiple machines or resources, it can be used much more efficiently.
As the physical resources are minimized, the reduced physical rack space means easier maintenance, smaller footprints, and reduced energy use. Data centers can use a significant amount of energy, so this last benefit can have a significant impact on operational expenses.
Desktops and servers can be cloned on existing machines, making provisioning and deployment faster and easier. Desktop virtualization through virtual desktop infrastructure (VDI) is probably the form of virtualization most commonly encountered by employees. VDI vendors include Citrix XenDesktop, VMware’s vCenter, and the open source Virtual Box.
In general, virtualization speeds deployment, setup, and maintenance of both servers and workstations, making life significantly easier for IT.
Hypervisor: The Technology Behind Virtual Infrastructure
Resources are allocated to virtual machines through a program called a hypervisor. A hypervisor is somewhat like an operating system for virtualized environments. It controls all the aspects of a virtualized machine, accessing and providing the resources of the actual physical devices. The term goes all the way back to IBM’s original mainframe memory sharing application in 1956.
Both of these platforms are supported with Veeam Backup and Replication 9.5, ensuring full availability.
The Need for VM Sprawl Management
Purchasing an actual server or workstation requires budget approvals, a procurement process, and inventory control, while a virtual machine seems essentially free. It can be created without a lot of negotiation, and eliminated just as easily.
As a result, IT practices surrounding VMs tend to be lax. Virtual machines tend to multiply without a lot of supervision. They are easily cloned, and in many cases abandoned without ever really being used. Not paying attention to these issues can create significant performance bottlenecks, when conflicting demands on CPUs and disk storage can slow things down significantly.
Virtual machines really are distinct environments. A proliferation of them can make security management difficult, with inconsistent or delayed patching. In addition, one great flexibility of virtual machines, the ability to undo, can cause security headaches, if a virtual machine reverts to a state before a recent security patch.
Virtual Infrastructure Management Best Practices
Cloud computing is a service that delivers shared computing resources, software, or data through the Internet. So virtualization is essential to the operation of cloud computing—all of the resources you access through the cloud are virtual. Those resources are managed by the cloud services provider, which certainly makes things easier for the user.
But many organizations still prefer to maintain their own internal resource, and optimize their provision with virtualization. The two can overlap, with a hybrid cloud solution.
The IT manager still needs to monitor and maintain the existing resources, keeping track of the physical machines, operating system, product licenses, and service level, while being ready to troubleshoot. A virtual infrastructure, while solving some resource allocation problems, does add a new level of management complexity, managing a more complex demand structure. The ease with which virtual machines can be created can lead to what some people call virtualization sprawl, or virtual machine sprawl, when more virtual machines have been created than can be effectively managed.
Vagrant is a free solution for managing the testing of applications, servers and configurations that is not tied to any specific virtualization platform, such as VirtualBox, VMware, or Hyper-V, but allows for the creation of workflows that work with any platform.
Managing Virtual Infrastructure
To keep operations efficient, minimize bottlenecks, troubleshoot, and keep security high requires proper virtual infrastructure management. You have to be able to discover and inventory every VM on your system, regardless of who created it, and perform ongoing performance monitoring.
The process of virtualization has moved ahead of simple, robust, and effective tools for managing it. The areas of performance analysis and troubleshooting show particular gaps, but other areas also have tools with an inadequate level of maturity.
So while virtualization vendors, including HP ConvergedSystem, VMware, and Microsoft’s Virtual Machine Manager, provide a built-in monitoring console, it still remains separate from the overall network, system, and application monitoring console, making it difficult to get an overall view to assign resources and troubleshoot.
With WhatsUp Gold's Virtual Monitoring you can monitor physical and virtual servers (hypervisors and guests) from one console. It all starts by automatically discovering, mapping and documenting physical servers and virtual machines, clusters, vCenter server, and VMware ESXi hosts and guests in your network. By harnessing VMware's built-in tools such as vCenter and vMotion, you can discover and monitor virtual machines, map them to corresponding physical machines, and track VMs as they move across physical machines in case of system failures.
And because it is also a WhatsUp Gold plug-in, you can easily trace the root cause of problems anywhere in your infrastructure. Get near-real time alerts as well as instant access to real-time and historical performance reports across devices, servers, applications and virtual resources in one single console so the entire IT team is aligned.