In its emerging phase, server virtualisation was mainly used with less critical applications in test-and-development environments. But as maturity has increased, more critical production applications are being deployed virtually, raising the stakes for creating the right architecture.
But even as virtualisation matures, storage remains a challenge: infrastructure and operations (I&O) teams have many vendor and technology choices to make, as the benefits of server virtualisation are dependent on each element of infrastructure that goes into the environment.
Today, server virtualisation is moving toward consolidation and simplification, with the following trends signifying a shift from “just making it work” to streamlining operations:
Server virtualisation adoption continues to accelerate. This year, the number of infrastructure teams using virtual server technology increased from 78 per cent to 91 per cent, continuing the trend of virtual server technology adoption for production applications. This adoption acceleration includes uses in more business-critical and performance-sensitive application environments.
Server virtualisation is being used for more critical applications. While web and infrastructure applications remain at the top of the list of applications running virtual server environments, databases and other more demanding apps have been increasingly used by developers this year. Microsoft SQL and Oracle databases and applications are increasing virtual deployments, indicating new willingness to put key OLTP applications in virtual server architectures.
VMware remains dominant, but alternatives have gained some traction. The number of organisations using VMware is down from 98 per cent to 93 per cent, but it still has the highest adoption numbers among hypervisor technologies. Other alternatives have gained popularity this year, with about 17 per cent of recent survey respondents using Citrix XenServer, placing it as the second most used hypervisor. Both Citrix and Microsoft are improving their products, and Microsoft’s pricing and bundling may allow the company to gain further traction against VMware in the coming year.
Managing storage for server virtualisation is hard, especially relative to efficiency. Effective capacity management was the greatest storage challenge last year, with 53 per cent of infrastructure professionals highlighting this as an issue. Maintaining high performance was only ranked a top challenge by 30 per cent of teams, down from 49 per cent in 2009. This decrease demonstrates a shift from creating a “good enough” environment for key applications, to fine-tuning a mature and well-understood virtual server environment.
As virtualisation is used for more critical workloads, I&O managers must increase the overall system reliability and make conservative decisions on the storage infrastructure behind these virtualisation tools. To achieve the benefits of consistency, infrastructures should be built around the virtualisation environment, making server and data protection key priorities for long-term benefits. It is important for I&O professionals to consider storage elements when selecting vendors, choosing a single vendor that also supports virtual servers, which will maintain consistency and eliminate unnecessary complexity. This means it will also be important to define requirements for storage equipment and test products, searching for one that will satisfy a single-solution environment.
By building a consistent storage environment for all virtual servers, enterprises can not only reduce complexity, but also allow for consistent data protection across a variety of applications and operating systems. And the value of this only increases as more critical apps are virtualised. But it will be important to consider backup, restoration, and business continuity issues for virtual applications and build levels of protection that keeps each virtual program protected within one of the standard deployment models. As maturity reaches its peak, infrastructure teams will also be able to explore alternative storage protocols —such as NFS or iSCSI — as an alternative to Fibre Channel. These options will offer simplicity and self-service capabilities with the cost advantages of standard Ethernet.
Andrew Reichman is a principal analyst at Forrester Research specialising in infrastructure and operations
Sometimes, the power of the mainframe is the most cost effective answer. Computing's Peter Gothard puts Computing's readers' questions on the future of the mainframe to IBM's Z13 expert Steven Dickens.
This Dummies white paper will help you better understand business process management (BPM)