Cloud Computing: A New Model for Delivering and Consuming IT Resources - Page 2

Wes Perdue

What Makes Clouds Different?

There are fundamental design differences that make clouds different from traditional on-premise data centers. Service providers typically consider their data centers to be value- or revenue-generating assets, while a traditional on-premise data center is commonly viewed as a cost center. This difference is one of the key reasons why the two environments are architected differently. Many of the large Internet data center companies share a philosophy on architecting a cloud environment, which is based on deploying low-cost, scalable, commoditized hardware with a software layer that glues the inexpensive hardware together to create a more scalable model.

Clouds are also based upon a self-service model, multi-tenancy enabled in part by virtualization, and a high level of autonomics and homogeneity. A traditional on-premise environment is typically a heterogeneous, shared-service model with less automation, and it manages assets independently for each user or application.  A cloud's deployment model is distributed, as opposed to decentralized with a client-server model. Another very important difference is that clouds are designed for failure instead of failover, which aligns well with their utilization of replication as the primary model for data protection due to its simplicity and massive scalability. 

Finally, open source software and cloud computing have a symbiotic relationship. While massive data centers built on low-cost commodity hardware and virtualized operating environments provide the technical foundation for global, cloud-based services, open source software enables global, service-based business models. These differences enable clouds to achieve lower service costs through greater resource sharing, greater economies of scale, greater levels of architectural standardization, greater process optimization and automation, and the ability to modify the usage of those resources much more quickly than traditional IT environments. 

Selecting the Right Storage Device for the Right Cloud Application

Applications drive the workload, reliability and performance needs of any system. Applications in cloud data centers run 24/7, have high workloads, and run in high vibration environments that require enterprise-class storage devices; e.g., hard drives (HDDs) and solid-state drives (SSDs).

Enterprise-class drives are designed for a multiple drive system environment such as servers and external storage systems that operate 24 hours a day, 7 days a week, 52 weeks a year. Non-enterprise class or personal storage drives are designed for a single-drive environment and have a much lower design usage, lower duty cycle and a simpler, lighter workload. To ensure success with demanding cloud applications, it is highly recommended to use enterprise-class drives that are designed for enterprise applications that run in a cloud data center environment.

Seagate's Unified Storage architecture is well aligned with the cloud architectural approach to increase simplicity, lower costs and achieve greater scalability. Seagate's approach reduces complexity and optimizes energy efficiency and performance footprint for the data center. By simplifying around a single interface and form factor, and upon a single foundation for security, Seagate's Unified Storage approach promotes sustainability and creates a foundation of powerful, yet simple enterprise-class storage solutions for the demanding and multiple array of different applications in the data centers. 

The Forecast: Growing Clouds

The cloud will continue to evolve and expand, driving more applications and more computing demand that will in turn drive more storage for the data centers, as well as more local storage.

However, the cloud will not wholly replace enterprise data centers. Many companies - primarily the small-to-mid-size businesses - will move from an internal IT model to an outsourced cloud model. Many larger enterprises, on the other hand, will have a hybrid model, utilizing the cloud for providing certain IT services while at the same time maintaining other internal IT services. Some may migrate and outsource their entire IT operations; however, many of them may choose to migrate their non-critical applications and data to the cloud but keep their critical applications and data internally, behind their firewalls. 

In 2010, there will be a huge growth in mindshare toward the cloud, with widespread cloud adoption taking place in the 2012-2013 timeframe. The degree to which companies will adopt cloud services for their diverse set of applications boils down to many factors; one of the more important of these is trust.  Earning this trust will take some time.

Add Comment      Leave a comment on this blog post

Post a comment





(Maximum characters: 1200). You have 1200 characters left.



Subscribe to our Newsletters

Sign up now and get the best business technology insights direct to your inbox.