Eliminating Data Center Vertigo with F5 and Microsoft

Without the proper feedback an automated data center can experience vertigo, leaving end-users dizzy and frustrated.

As organizations continue to virtualize and automate the data center in their quest to liberate themselves and their users from the physical bonds that have kept them tied to the data center floor they are necessarily moving “up the stack” and running into a profoundly important question: how do I enable IT as a Service?

Virtualizing compute, network, and storage resources is just the first step. Once those are virtualized, they must be managed. Once they’re managed, the next layer of the stack needs to be addressed. At the top of that stack is IT as a Service and it, too, must be managed. Business users cannot leverage the automation and efficiency provided by cloud computing and a dynamic infrastructure without an interface and it is here where the killer application for private cloud computing resides. But without the proper feedback from the relevant components, that efficiency can be quickly lost as resources are provisioned in a volatile, unpredictable manner to try to regain balance in the data center.

A STRONG FOUNDATION

Ultimately, business users and internal IT consumers want push-button provisioning capabilities in the data center and IT must enable that through a strong foundation of virtualization, infrastructure, and data center management technologies.

That foundation must provide for key cloud capabilities: availability, elasticity, automation, and self-service. These four capabilities are intimately related to each other and any one without the other leaves a hole through which resources and efficiency can leak.

By offloading the manual operations and provisioning of individual workloads (applications) from core IT with its Self-Service Portal offering, Microsoft provides the foundation necessary to enable an architecture capable of supporting all four key capabilities required of an enterprise cloud deployment. Recognizing that network and storage resources must work in concert with compute resources, Microsoft has been hard at work enabling the means by which the infrastructure components comprising the foundations of enterprise cloud computing can be integrated, provisioned, and managed by core IT. All resources must be balanced across application and user need and demand, and to do so requires a unified management system that engenders collaboration across the infrastructure and provides the consumers of those resources with the means to manage them effectively.

The two-way integration between F5 and Microsoft in terms of management and run-time dynamics is important because a private cloud computing deployment is more than just server virtualization. It’s a combination of all the technologies required to self-service provision and manage an application; an application that requires more than just a deployment platform. Applications require storage, and a network, and integration with other applications and systems. Making an application “elastic” requires a careful balance of provisioning not just a virtual server but an entire ecosystem of storage, network, and application network resources. It is not enough to simply provision a virtual server; the goal is to provision and manage and deliver applications as elastic, dynamic data center components. To accomplish this task requires visibility into the network and the application, which ensures the infrastructure is able to adapt dynamically to changing workloads and properly balance costs and performance with the consumption of network and compute resources.

BALANCE REQUIRES FEEDBACK

Anyone who has tried to balance on anything when their hearing is for some reason impaired knows that its hard to do. If the feedback we need is impeded or completely cut off, we lose our balance. The same is true for all three data center layers. Feedback is imperative to balancing resources, applications, and ensuring that the consumers of IT as a Service have the data necessary to make informed decisions as they begin to leverage IT as a Service and manage their own resources. Without such feedback, an application (and ultimately the entire data center) can experience a bad case of vertigo, its performance and availability going up and down and sideways, leaving users and customers dizzy and frustrated.

Microsoft recognizes that the IT infrastructure as a service model is necessary as a means to enable this balance as IT moves to deliver and manage its resources as a service. By integrating F5’s BIG-IP application delivery controller (ADC) with Microsoft System Center, the network becomes visible and application load optimization policies can be automatically managed by System Center based on real-time conditions within the data center. This allows System Center to automatically add or remove application virtual machine instances as application load thresholds are reached and F5’s BIG-IP to automatically adjust its load balancing to accommodate. The net effect is that the application is always running at optimal load levels for users maximizing performance and eliminating waste.

F5 and Microsoft Virtualization technology work together to create a dynamically provisioned datacenter that uses resources more efficiently, improves application performance, and delivers a better user experience. Client traffic flows through the F5 BIG-IP Local Traffic Manager (LTM), which makes intelligent routing decisions to maintain the best performance for all users. When integrated with System Center Operations Manager through the F5 management pack SCOM is able to monitor traffic load levels for increases or decreases that affect performance and to take appropriate action to maintain high performance and availability.

The integration between F5 and Microsoft enables administrators to automate the reaction of the delivery infrastructure based on a holistic feedback loop to meet the demands of the consumer and adjust as defined by preconfigured policies the compute and network resources necessary to meet desired service levels. By balancing resource consumption with needs automatically and based on real-time demand, a joint F5-Microsoft infrastructure as a service solution can eliminate availability vertigo and keep applications – and their users – balanced.

For additional details on Microsoft's SSP offering, visit their SSP blog.