How To Future-Proof Your Cloud Deployment

Computers & TechnologyTechnology

  • Author Akhilesh Rajput
  • Published April 19, 2016
  • Word count 1,176

Cloud computing, a type of computing that relies on sharing computing resources rather than local servers to manage applications is an on-demand service model for IT provisioning and is based on virtualization and distributed computing technologies. Cloud computing has radically changed software development and the way enterprises use data and information. It has become a mainstay of the current IT landscape, with the market research firm, IHS, predicting that £142 billion will be spent by enterprises globally on infrastructure and cloud-related services by 2017.

However, when considering an enterprise cloud solution, it is important to note that rather than being a single solution, it is a combination of technologies includingvirtualization, networking andstorage solutions, and hardware - working together, with each of them having their own unique characteristics. These components, and their characteristics, should be adequately understood before enterprises adopt cloud solutions. There are significant benefits to using cloud solutions. The main advantages are rapid go-to-market, business performance resourcing, business agility,and cost reduction. Many companies across the business spectrum are realizing the benefits of cloud and are migrating or will migrate to a private, public or hybrid cloud environment. While there are definite benefits to using cloud solutions, proper planning and diligence can ensure that a company’s cloud investments withstand future change or become future-proof.The sections below outline steps to optimize cloud investments.

Data Center Optimization

One of the key steps to perform before a company considers cloud migration is to ensure that its data centers are in order and automate its data center virtualization processes. Data center virtualization and cloud computing are often thought to be the same. However, while virtualization is a fundamental cloud computing element that helps deliver on the value of cloud computing, cloud computing refers to the delivery of shared computing resources, software or data, on-demand and through the Internet. Virtualizing the data center can significantly ease the transition from the data center to the cloud.

Planning and Analysis

The type of cloud environment that a company adopts is largely determined by the corporate demand and existing IT resources. These cloud environments can vary greatly since every company’s needs are unique. The IT resources a company adopts should ideally be utilized fully, and for this, planning is critical. Questions such as what is going to be accessed and by whom, how users will connect to the cloud environment, how many users the environment will need to support, and how growth can be accommodated need to be thoroughly analyzed.

Cloud Deployment Strategy

The cloud deployment strategy will differ based on the organization’s needs and factors such as which cloud environment (or combination) - whether public, community, private, or hybrid clouds has been chosen. To successfully deploy a cloud environment, the workload type, WAN/LAN bandwidth, current and future resource needs, and user training and accessibility need to be understood in detail. The workload depends on the type and amount of data being deployed to the end user, and will vary depending on whether a small application or full-blown virtual desktops are being deployed. Determining the workload can help in choosing the best cloud environment. An appropriate WAN/LAN bandwidth can mean the difference between a robust cloud environment and one that suffers from detrimental latency issues arising from WAN/LAN congestion. WAN optimization technologies can also be used for efficient bandwidth allocation. Properly understanding a company’s needs, both current and future, will help the company respond well to future growth. In addition, an enterprise with a cloud environment needs to address difficulties that end-users may face when using its services through the cloud. Planning for and preparing adequate material to explain the cloud environment and functionality to the end usercan mean the difference between the success and failure of a company’s cloud-based offerings. Increasingly more users are looking at the quality of a product’ssupport, such as easy-to-follow training manuals, webinars, and documentation for easy cloud adoption.

Understanding Application Workloads

There are primarily two types of application workloads: traditional scale-up and cloud-native scale-out application workloads. These two workloads have vastly different characteristics, which need to be understood by enterprises looking to build a cloud environment. Most of today’s enterprise applications, such as SAP ERP, Oracle database apps, and Microsoft Exchange, which reside in data centers come under the traditional scale-up category. The workloads are generally client-server or n-tier applications that run on a single server or a group of servers and databases. To scale-up traditional apps, the preferred method is typically to increase the size of the application and database infrastructure, and therefore, the ability to scale-up for traditional apps is somewhat limited by the workload architecture. The new generation of apps generally associated with cloud computing, such as gaming and mobile apps,Big Data, and social apps, fall into the cloud-native scale-out application workloads category. These workload types require a substantial amount of dynamic scaling and elasticitythat would be very expensive or even impossible to achieve with traditional datacenter architectures.

With the development of cloud computing, computing power began to be provided by multiple commodity grade computing, networking and storage nodes rather than increasingly larger and more expensive servers. The load generated by millions of users is cost-effectively handled by internet companies like Google, Facebook, and Amazon by running multiple application servers in parallel, utilizing significant caching methods, and replicating data to numerous traditional or distributed database servers. Traditional scale-up and cloud-native scale-out application workloads have many other differences in addition to those stated above and a company looking to deploy a cloud solution has to seriously consider the two options to determine the solution that best fits its needs.A relatively new cloud offering, Citrix’s Cloud Platform offering, which is powered by Apache Cloud Stack, is an application-centric cloud that could indeed be used for both traditional scale-up and cloud-native scale-out application workloads would certainly be valuable considering future growth and expansion.

Reliability, Performance, and Security

The public internet serves as the default for accessing cloud applications and is generally adequate for consumer applications. However, for enterprises with critical applications in the cloud, a private network should be considered for reliability, performance, and security.Selecting a cloud solution that offers global reach is also important for companies with global operations or those looking to do so. Providing private, secure access to corporate resources from any location can be critical to an enterprise’s operations. The flexibility to increase bandwidth when necessary should also be considered when choosing a cloud solution. Companies that require an adequate bandwidth at all times without having to pay for idle capacity should consider on-demand, burstable bandwidth models.

Cloud computing is still developing and will continue to evolve. Further, the needs of a company change constantly, and therefore, there is no one single correct cloud solution - each has its pros and cons. The company whose cloud investments continue to generate returns well into the future will most likely be the one that assesses its needs in depth and chooses a cloud offering or combination that best fulfills its needs.

This article has been viewed 1,632 times.

Rate article

Article comments

There are no posted comments.

Related articles