The Cloud as a Datacenter Extension
Article by Karan Kirpalani
- Filed under:
- Cloud Computing
It's been said that cloud computing is the new computing paradigm − that before too long, enterprises will be no more willing to build, maintain and operate their own data environments anymore than they would their own electrical grid.
But while many organizations may one day turn to the all-cloud datacenter, others simply have invested too much money building data infrastructure over the years to cast it away just because something new has come along. Still, that doesn't mean the cloud is of no value to established data environments. In fact, the exact opposite is true: the cloud is proving to be an effective tool in supplementing traditional infrastructure as enterprises come to grips with the Big Data needs of newly mobile users and modern database and analytics platforms.
But exactly how the cloud can best be of service to standing enterprise infrastructure depends on several factors. Certainly, the nature of that infrastructure has a lot to say about what kind of cloud model is most appropriate, as do the needs of end users and IT departments.
For instance, an organization that has built even a moderately pervasive virtual environment would most likely benefit from an enterprise-grade IaaS (Infrastructure as a Service) architecture. At this level, additional virtual resources can be created and integrated into existing environments with very little up-front cost. Users simply pay for the resources they need at a given time and then raise or lower their consumption as data needs change. Cloud resources remain under complete control of the enterprise, as if they were merely an extension of their own virtual infrastructure, while the provider takes care of managing and upgrading the physical underpinnings of the cloud environment.
In many cases, cloud services are delivered using shared resources; that is, the infrastructure used by one enterprise is also available to others, with sophisticated multi-tenancy and security software preventing the comingling of data. Some organizations, however, are opting for private cloud services. The common perception of the private cloud is one in which cloud architectures are built within the enterprise's own data infrastructure, but this is not the only way it can be done. A third-party cloud provider can carve out dedicated resources within its datacenter, providing an external private cloud that is shared with no other clients. This approach is generally available at a premium over standard cloud infrastructure, but many enterprises find the cost is reasonable for the peace-of-mind it provides in support of higher-value data loads.
Still other organizations have moved beyond simple virtualization and are building their own internal clouds. In this situation, a hybrid infrastructure offers the best balance of low cost and high flexibility and efficiency. The hybrid cloud combines both the internal and external cloud architectures through either standard or proprietary platforms in such a way that both data and applications view them as a single, integrated environment. Enterprise policies, network topologies and other features can easily pass from one side to the other, maintaining a highly fluid, highly scalable environment. The hybrid cloud is seen as a key component in what some are calling "cloud bursting," which allows enterprises to quickly and easily provision additional resources should data loads suddenly spike.
With all these options available, the cloud is the most flexible data environment ever devised. It provides a tremendous amount of agility when tailoring infrastructure to meet data needs, even as those needs are in a constant state of flux.
No matter what challenges confront the enterprise going forward, there is a good chance that there is a cloud model ready to meet it.