Q&A: Managing Multiple Cloud Environments
First it was just "the cloud." Then came public clouds versus private clouds, and, of course, hybrid clouds. How is IT coping, and how do clouds fit in with other trends, such as the distributed data center?
Once a simple proposition, cloud computing has grown in popularity and complexity. Now IT must manage multiple cloud environments -- public, private, and hybrid -- all within a distributed data center. For perspective on the growth of cloud and how IT is taking advantage of the technology, we spoke to Nand Mulchandani, co-founder and CEO, ScaleXtreme, a company that specializes in server management solutions including cloud-based server automation.
Enterprise Strategies: Why is there such a focus on managing multiple cloud environments in 2012?
Nand Mulchandani: Cloud computing is no longer in its infancy, and companies are realizing they must reevaluate their business models to include the cloud. They're finding new ways to use this remarkable technology but tripping up over how to handle the sudden spike in infrastructure complexity. They have remarkable new power from public cloud providers such as Amazon, Rackspace, Bluelock, Terremark, and others but they are having trouble wielding in an effective way -- especially when they try to combine those public cloud instances with private cloud computing and their legacy infrastructure of physical servers. Getting the most out of this complex environment means focusing on management.
How and why are companies deciding to adopt the cloud in parallel to their on-premise and legacy systems?
You just can't do everything in the cloud -- at least not yet -- and then there's the fact that most companies have expensive legacy systems in place that they're not ready to replace. They also have lingering questions of cloud security, availability, and cost.
The truth is that most companies will maintain some level of physical server infrastructure for years to come. It will handle their most sensitive data or their most critical applications. Smart systems architects are using the public cloud less for the old applications that physical servers can handle well and more for exciting new projects that wouldn't have been possible without cloud technology.
How do IT departments today typically manage their server environments? What deficiencies are there in this approach?
Enterprise IT departments are constantly faced with complex data integration, security, and system reliability issues. To date, IT departments and sys admins are on-call around the clock to ensure that any unexpected system failures or bugs do not jeopardize the integrity and productivity of its servers and data centers.
Often times, the enterprise has a combination of traditional on-premise servers, virtual machines, and cloud solutions, making uniform monitoring, provisioning, and management of the entire server environment laborious and very time-consuming. Vendors are responding to these current challenges. For example, ScaleXtreme, the company I work for, created an all-in-one solution that allows sys admins to automate all of their routine server management tasks through a single framework. In addition to ease of management, we realize that data security is a top priority so we recently launched a free cloud-based patch management feature for our systems management solutions for Microsoft Windows and Linux Systems in physical-, virtual-, and public-cloud servers.
How does the public cloud fit into the concept of a distributed data center, and what benefits does a company get by thinking holistically about its cloud-, virtual-, and physical-server deployments?
Cloud computing is undeniably great and a major advancement in the fundamental way people will build and use computers, but that doesn't mean they'll be ditching their physical, on-premise data center equipment. The cloud is another evolutionary branch of the computing tree -- another part of the modern, distributed data center. As the cloud becomes a crucial and integral part of most businesses, it will be important to harness the ability to manage all of your data in one place, easily and cost-effectively.
What is the importance of an enterprise diversifying their compute portfolio? Are there any drawbacks to such diversification -- for example, is it more difficult to manage?
A diversified computing environment allows systems architects to really get creative. Our customers tell us that they opted for some combination of physical servers, virtual machines, and public-cloud instances to do something new and exciting that they wouldn't have been able to do even as recently as five years earlier. That can be on-demand provisioning, usage-based bursting, agile development instantiation, or any number of other impressive and imaginative new use cases.
However, this ability to do new and interesting things comes at the cost of additional complexity. IT managers, right up to the CIO, may not know why they're paying for various computing resources. There's little clarity and great difficulty managing this highly heterogeneous and rapidly changing environment.
What are some of the common misconceptions about cloud computing? How will the distributed data center allow IT administrators to better manage their systems?
The biggest misconception people have about the cloud is that it will save them money. Renting servers from Amazon, Rackspace, or another provider can save some customers money, but it isn't cheaper for everyone. Customers with constant workloads and predictable capacity may find it less expensive to buy and manage their own servers. There's a learning curve to come up and there can be a significant cost of complexity going to the cloud.
Our customers are finding that public cloud computing can dramatically improve their operations and reduce costs when combined with traditional physical servers. That gives them the opportunity to try new things and the capacity to take on rapidly growing compute loads while preserving their underlying infrastructure efficiencies.
In an environment when companies are trying to cut costs on their IT departments, what benefits could the distributed data center realistically provide?
The old model of computing focused on preventing catastrophic failure to the exclusion of all else. IT managers put all their eggs in one basket and maniacally watched that basket. When budgets permitted, they installed a second basket underneath the first.
The distributed data center has three major components: physical servers, virtualized servers, and cloud servers. Each has its own place in the enterprise and has an optimal set of associated applications. The ability to make tradeoffs is the powerful root of the distributed data center and an increasingly important function of IT decision makers as tech becomes more embedded into every aspect of business.
What best practices do you recommend to actually achieve these benefits?
Start by thinking carefully about what you really need your computer systems to do, then determine if adding virtualization and public cloud computing might make sense. Invest the time and effort to figure out how you'll manage the additional complexity of taking your business to the public cloud, then experiment with various cloud providers and configurations until you find something that works well for your organization.
If you've done a good job setting up your systems management, you'll be able to test and improve your overall IT infrastructure easily over time.
What products or services does ScaleXtreme offer related to our discussion?
ScaleXtreme products provide different levels of server management capabilities. ScaleXtreme Xpress is a free cloud management solution for the single user who needs end-to-end life cycle automation and management for servers in a single environment. New customers or upgraders can use ScaleXtreme Xpert; at $5 per server per month, it helps solve the biggest server management problems associated with Web 2.0 organizations. Our newest product, Xtreme, enables the highest levels of security, efficiency, and flexibility; it's $15 per month per server.
These solutions reduce time spent on manual, low-value repetitive tasks. All three products feature Dynamic Server Assembly that helps systems administrators take advantage of cloud computing by creating highly structured servers. Users can model a server once, quickly ramp up multiple instances, and run them anywhere.