October 27, 2008
As the IT world continues its march toward a service-oriented future (read: cloud computing), the spotlight continues to shine on aspects like virtualization, computing and billing models -- and rightfully so. Lost in the mix, though, or perhaps taken for granted, are automation technologies, which some believe are the foundation for any legitimate cloud computing infrastructure.
Take, for example, Venkat Devraj, co-founder and CTO of datacenter automation provider Stratavia, who has been helping customers like Xcel Energy, Visa and several financial services firms enable their internal cloud infrastructures. Devraj sees any cloud offering as being a stool with three legs: virtualization, SOA and automation, respectively. Whether it is storage virtualization, server virtualization or whatever, the level of abstraction virtualization provides is critical to the notion of a cloud. Concepts of SOA, he says, provide the dynamism and agility needed to deliver services on demand.
The third leg, datacenter automation, brings the capability of real-time, or on-demand, infrastructure management, says Devraj. It does so by allowing resources to be managed effectively behind the scenes as they are provisioned, and during post-provisioning activities like configuration management, change management, incident management and maintenance, among others. “I’m seeing that datacenter automation is kind of the glue of the third leg of the stool that brings together concepts of virtualization, SOA and so on to bring an abstract offering to the end-user community or to the business-user community,” he says.
He adds that while some vendors and critics dismiss cloud computing as a rehash of old ideas and computing models, that criticism just is not accurate. “The true nature of the cloud is resources behaving [autonomously], resources behaving dynamically, with SOA, with virtualization, with datacenter automation, and these are concepts that didn’t have mainstream status even two years ago, even a year ago, to be honest,” Devraj says.
Robert Gardos, CEO of database automation provider GridApp Systems, agrees that automation is an integral element of any cloud infrastructure. Whether it takes place on the user side (managing hosted machines) or the provider side (updating infrastructure and provisioning VMs), “Automation is the building block of any self-service infrastructure,” he says, “and I don’t see how the cloud could possibly work without being self-service.”
As for his company’s sweet spot with the database, Gardos says that companies will need software to maintain their backend database management systems as they move into an on-demand delivery model. He says GridApp helping a number of its banking customers to “rapidly deploy technologies like Oracle RAC, empowering the users to quickly scale those environments or relinquish capacity, and then facilitate some sort of internal billing process.”
Richard Muirhead, CEO of Tideway Systems, believes automation enables cloud or hyper-scale infrastructures by letting companies understand the complexities and dependencies that affect application or service performance -- especially in such large, multi-faceted datacenters. “You can’t deliver high-level services … that require the effective operation of a bunch of different technologies … reliably and economically to your internal customers unless you understand them and you can track them and you can monitor them, and you define even what they should be and can assess and police them,” he says.
Muirhead says Tideway’s suite of products, including its flagship offering, Foundation, addresses this problem by automatically mapping applications to the underlying physical and virtual infrastructures. Its new Configipedia offering (which Muirhead describes as a user-centric “encyclopedia of IT configuration”) also plays a role by helping IT teams find bad configurations that will have a negative impact on performance or availability, or, conversely, search for good configurations that fit a business’s particular needs. In virtualized environments, he says, basic automation products can solve template management issues, but you also have to consider configurations around VMs up and down the stack.
Being able to understand your environment “is going to be absolutely central to making cloud computing work, especially when you’ve got to have a tight understanding of SLAs, if you are going to take your applications and throw it over the firewall into someone else’s environment and pay them for the privilege in some way,” he explains. “That can only happen through consistency of the configuration and automation and industrialization, if you will, of the whole process.”
Devraj, too, sees automation playing a big role in optimizing cloud environments by managing this growing complexity. “Previously, for example, you could push out a new application release very easily; it would be contained in a certain environment,” he says. “But now, with a lot of dependencies up and down the stack … even a simple change management task can break an entire stack.”
Andi Mann, research director at Enterprise Management Associates, has the following to say about datacenter automation: “Once you get to a certain size -- and that’s not a very big size, either -- it becomes absolutely critical.” You simply cannot manage a large-scale datacenter manually and efficiently, he says, and “[i]f you built out your datacenter as your company grows simply by throwing more people at it, you would literally end up at the point where you can’t fit enough people in your datacenter.”
He adds that he sees a nexus of virtualization, cloud computing and datacenter automation, and enterprises need to realize that these three aspects must be managed together. EMA research shows that a typical enterprise doing virtualization has to deal with 11 different platforms from several different vendors on just the virtualization layer, and the complexity and interdependency among the various tools and layers does not lend itself to easy manual management. “It’s very, very difficult, if not impossible, to provide computing as a service when you’re handling it all manually,” he says. “It just does not make sense.”
“Automation forms the basis for any advanced computing techniques,” he explains. “If you want to go to cloud computing without any automation, what that fundamentally means is you don’t really have any repeatable, reusable processes. What that means is you’re trying to outsource something that you don’t really have any handle on yourself.”
Users Starting to See the Connection
Stratavia’s Devraj sees cloud computing providers falling into three main categories: heavy-duty providers like Amazon, Google and, now Microsoft; old-school hosting providers trying to enable cloud offerings; and, finally, individual organizations providing private cloud resources to their internal customers. He says providers in the first category don’t trust anything they don’t build themselves, so trying to sell to them is “futile.” Where Stratavia does the most business, he says, is in the third category, where Fortune 1000 CIOs want to build public cloud service into their own datacenters. Whether they build it themselves or buy it commercially, Devraj reiterates that automation is key for a sustainable, scalable cloud business model.
This type of engagement mirrors what Devraj is seeing in the automation marketplace overall. Awareness of datacenter automation is up, he says, but still needs to improve, “and the only way it’ll come about is through the glamour and sexiness of architectures like cloud, which will force them to build it up first in a certain manner, watch it being broken, watch it being abused, and then getting all the ingredients.”
GridApp’s Gardos believes automation technologies currently are being utilized mainly to ensure repeatability and efficiency within datacenters, but acknowledges that automation as a precursor to on-demand delivery models definitely is on customers’ radars for the next few years. In discussions with CTOs and CIOs, he says, “The biggest question is: What happens when it’s a rainy day?” This mindset means it will be a while before mainstream, mission-critical applications move into the cloud, he adds, but it will happen because “it’s foolish not to do that.” When it does happen, GridApp will be there to manage the database automation aspect.
Tideway’s Muirhead says that although users and software vendors are getting more sophisticated in understanding the challenges surrounding cloud computing, there still are many variants that need to be addressed -- some as simple as which platform to choose -- all of which need to handle the automation problem. Regardless of the obstacles left to overcome, Muirhead says business executives see what external cloud platforms can offer, forcing CIOs to raise the bar internally to compete. “The fact that the concepts are being defined is accelerating the pace at which your Wachovias, your JPMorgans and all these guys are looking at ‘How do I define an infrastructure template or pattern, for example, that I can repeatedly deploy and offer up as a service -- essentially a cloud service -- to application development teams internally?’”
May 16, 2013 |
When it comes to cloud, long distances mean unacceptably high latencies. Researchers from the University of Bonn in Germany examined those latency issues of doing CFD modeling in the cloud by utilizing a common CFD and its utilization in HPC instance types including both CPU and GPU cores of Amazon EC2.
May 10, 2013 |
Australian visual effects company, Animal Logic, is considering a move to the public cloud.
May 10, 2013 |
Program provides cash awards up to $10,000 for the best open-source end-user applications deployed on 100G network.
May 08, 2013 |
For engineers looking to leverage high-performance computing, the accessibility of a cloud-based approach is a powerful draw, but there are costs that may not be readily apparent.
05/10/2013 | Cleversafe, Cray, DDN, NetApp, & Panasas | From Wall Street to Hollywood, drug discovery to homeland security, companies and organizations of all sizes and stripes are coming face to face with the challenges – and opportunities – afforded by Big Data. Before anyone can utilize these extraordinary data repositories, however, they must first harness and manage their data stores, and do so utilizing technologies that underscore affordability, security, and scalability.
04/02/2012 | AMD | Developers today are just beginning to explore the potential of heterogeneous computing, but the potential for this new paradigm is huge. This brief article reviews how the technology might impact a range of application development areas, including client experiences and cloud-based data management. As platforms like OpenCL continue to evolve, the benefits of heterogeneous computing will become even more accessible. Use this quick article to jump-start your own thinking on heterogeneous computing.