With more than ten years of public cloud infrastructure R&D and customization experience, Inspur is a trusted partner for your business success.
There is no lack of interest in cloudbased solutions to address many of the challenges facing IT today. Forecasts for the cloud, whether for the purchase of public cloud services or for infrastructure to build a private cloud, vary. They differ, depending on a number of factors and whether hardware, software or services are involved. However, annual growth rates of 18% to 20% are typical, which is impressive for IT markets.
While interest in cloud is high, the decisions as to how and when to move to the cloud, and in what form, are still being made in many organizations. As with any decision involving IT architectures and approaches, the cloud requires an in-depth understanding of stakeholder needs, existing IT assets and how those assets are being used. The business must also have the ability and willingness to embrace and support the transformations involved.
Some of the most serious considerations organizations face in implementing a cloud involve security and control. Traditional data centers have their limitations, but since they reside within the company, IT can exercise strict control over them. Moving to the cloud can mean giving up some or all of this control. Such a move introduces security concerns, real and perceived, that can slow down cloud deployments or even stop them completely. These concerns are especially evident in cases where data is sensitive and workloads are critical.
Organizations have a choice of cloud options: hosted, private and hybrid. An outside vendor owns and hosts the hosted clouds, while private clouds reside within the company and are under its full control. Hybrid clouds combine hosted and private clouds: Well-chosen applications and data are deployed to 1 or more public clouds, while sensitive data and mission-critical applications are relegated to their private cloud. There is full integration between the 2 environments.
While public cloud providers claim to provide highly secure multitenant environments, security concerns have moved many organizations to begin with a private cloud. For some companies that have done so, their experience and success have given them the confidence and know-how to consider deploying some workloads to a public cloud.
Likewise, meeting the stakeholder demands noted earlier means modernizing the data center, which involves steps that are necessary to take full advantage of private cloud benefits. A full private cloud is a logical extension of the work already being done to make IT “service-driven.”
Private Cloud Infrastructure and Challenges
Competitive pressures have been pushing organizations to improve IT’s capability to support the activities of internal stakeholders that deliver the highest returns to the business.
For instance, line-of-business managers must access the business applications and data they need on demand. Developers require the ability to provision resources to create, qualify and test new business-strategic applications in a timely manner.
To support the varied but stringent requirements of these and other end users, IT’s emphasis must move away from maintaining applications tied to siloed infrastructures to a more services-oriented model.
Achieving the transformation required to make this services orientation a reality can result in significant value, but there are challenges to take into account along the way:
More Efficient Asset Utilization.
Significant savings can be achieved by consolidating resources, which results in those resources being used more efficiently and, ultimately, reducing unnecessary purchases. Doing so requires skills, technologies and, in some cases, new organizational structure or alignment.
If proper planning hasn’t taken place, transformation can cause significant disruption to operations as new technologies and platforms are brought online. New processes and skills must be developed and learned. The new, modernized infrastructure must therefore be created and deployed quickly, with existing assets leveraged as much as possible to minimize downtime and any possibility of lost business.
Managing Application Complexity.
Often, applications are numerous, built using differing technologies, and span a wide range of workload requirements. Some are virtualized, while others are not. Gaining a thorough understanding of what applications exist and what is being used today, and the storage, runtime and integration requirements for each is critical. Once those applications are migrated to a newly trans-formed data center architecture, the results can yield significant benefits: Organizations see greater agility, faster time to value, and more optimal resource usage.
Transforming to a flexibility-on-demand architecture can require new tools and capabilities as well as organizational and operational process refinement. Automating the processes of provisioning, deprovisioning and managing resources overall is critical to achieving the efficiency, speed, performance and cost benefits of a transformed data center. Doing so effectively can make IT more productive by freeing up resources for more strategic tasks that benefit the business.
The Steps to Private Cloud
Whether a full private cloud implementation is your goal, or you want to transform and modernize all or parts of your data center infrastructure, your success depends on the approach you take. Using an organized, multistep process that starts with building a cost-effective, efficient foundation can help you fully leverage the benefits of modernization and a services-driven environment.
Our methodology to transform traditional IT to a private cloud involves 4 major steps that build on each other (see Figure 1). The goal is to create a fully integrated and efficiently managed pool of IT services capable of providing stakeholders with the resources and service levels they require:
The 1st step involves reducing your IT footprint to lower overall costs by reducing capital expenditure (capex), lowering space and utility costs, and lessening the administrative resources required. The goal is to architect 1 infrastructure platform for all data and workloads, which increases efficiency by minimizing costs and maximizing utility. For some, this consolidation involves only storage platforms, for others, compute resources, and in some cases, it takes the form of a “converged” infrastructure of all resources. In all cases, the resulting consolidated infrastructure is more efficient and cost-effective. If the consolidation activity is able to leverage your existing resources without any significant “rip and replace,” these benefits will be particularly evident.
The 2nd step allows you to extract greater value and to become more efficient. You can create a single virtual pool of all resources, including file, block and object storage, servers and applications, with the ability to manage across all of these resources. Resources in this pool can then be dynamically provisioned on demand (see Figure 2). The resulting flexibility can mean more optimized usage because resources are no longer tied to legacy application or platform “silos” that remain unused when they can be leveraged somewhere else.
Ideally, this virtualization can be applied to single, converged infrastructure where both storage and compute resources work together to get tasks done. Then, processes must be revised to take advantage of the new virtualization capabilities. The IT staff must be trained and business stakeholders must be made aware of the new processes.
The 3rd step is to automate your IT management and administrative processes to enhance your ability to meet committed service levels while making IT and stakeholders more productive. When applied to an infrastructure composed of integrated pools of compute and storage resources, and a complex mix of applications, automation can ease managing and provisioning tasks. Automation can deliver services with greater speed and reliability. As the new processes have been streamlined, automating them creates a highly efficient IT organization. With this type of approach, policies can be effectively enforced, and risks associated with tasks prone to human error can be reduced.