October 31, 2008
If you accept the premise that the majority of computing will move into the cloud, you might wonder what different types of clouds there might be. Serguei Beloussov, CEO of software vendor Parallels, foresees five types, all grounded in the present and none promising anything like free hover-bikes or time machines.
The first three types fall into what Beloussov calls platform clouds. “That’s the Google cloud and now the Microsoft cloud. Those two are going for the whole world. They have invested millions and millions in their datacenters, and they are providing developers with tools to build on top of their infrastructure and services,” he says.
Then there is the third type, in which vendors provide a cloud-based platform for their customers. Beloussov cites IBM, for example, as building clouds to serve high-end enterprises and enterprise applications. On the other end of the spectrum is Apple, which has built a cloud for consumers with Mobile.Me. Facebook, he says, is a social cloud. And, of course, there is Amazon with EC2 and S3.
Another category is what Beloussov calls “services clouds.” “These would include type 4 clouds, which are the ‘channel’ clouds of service providers, ISPs, telcos, Web hosts,” he explains. “And then there are type 5 clouds, which are in-house clouds that will be built mainly by large companies, to serve affiliates and subsidiaries or different departments.”
“Everybody from consumers to enterprises will eventually move into one of those types of clouds,” he says.
But Beloussov’s eye is mainly on types four and five. “We believe people will be moving into those different clouds, and our business is to provide virtualization, automation, and management for type 4 and 5 clouds. Virtualization and consolidation will lead to clouds with millions of instances of virtual machines, applications, servers and containers, and that will mean a need for automation and extreme efficiency.”
If you’ve ever been on the Internet, there’s a good chance you’ve brushed up against one of Parallels’ products. The company’s broad lineup includes server virtualization, automation and management tools used by some of the world’s major hosting services. Beloussov says Parallels is “No. 1 in a few areas: desktop virtualization, OS-level virtualization and server automation control panels.” (Those rankings come by way of Parallels Desktop for Mac, Parallels Virtuozzo Containers and Parallels Plesk Control Panels, respectively, he says.)
Virtualization Two Ways in One
“There is no silver bullet” when it comes to server virtualization solutions, Beloussov says, so Parallels offers two schemes: hypervisor and container. “In the hypervisor approach, made popular by VMware, the software emulates hardware, and you have the OS running inside the virtualized machines. Our Parallels Server product takes this approach,” he says, “A different approach is OS virtualization or container virtualization. This is the approach of our Virtuozzo Containers product. The software does not imitate the hardware, but rather virtualizes an operating system itself. One fundamental difference is that with the hypervisor approach, customers end up with multiple copies of the operating system on the physical machine, whereas in the container approach, there’s only one OS per physical machine.”
The biggest advantage of using containers is scalability, Beloussov explains. “Some of our customers run 500 containers on a single machine. That’s not realistically possible with a hypervisor approach. Virtuozzo Containers allows you to partition a physical machine into hundreds or even thousands of containers. This type of density isn’t possible with hypervisors because you have to load a complete copy of an OS in the virtual machine. This complete copy of an OS takes resources. You have to have an additional 256 megabytes just for the operating system itself. In the case of a container, there’s no OS to load because the host operating system is virtualized.”
For something like test and development, though, you want to be able to run different operating systems on the same hardware. “You can only do that with the hypervisor approach,” he says. “Legacy server consolidation, too, would be possible only with a hypervisor system.”
The goal is to have the right technology for the job, he says, and toward that ideal Parallels is planning to combine the two approaches into one product. “You’ll be able to integrate containers or full hypervisor virtual machines on a server,” Beloussov says. This “integrated virtualization offering” is slated to be available next year.
Automation tools are the other part of Parallels’ plan to be part of the service clouds ahead. In order to handle service plans for millions of subscribers, providers will need completely automated management tools, he says. In the cloud, automation means provisioning of applications on a large scale, and it means handling workflow -- internal and external. Usage charging must also be automated, he adds.
Beloussov believes Parallels has the core capabilities to meet these needs in the cloud now. The company’s catalog covers systems automation (provisioning, application management), business automation (account management, billing), and control panels used by very large hosting providers to manage their sites. Says Beloussov: “We pretty much automate everything: provisioning, Web infrastructure, middleware, physical and virtual infrastructure, as well as business processes.”
Short Timeline for Cloud Ubiquity
As for the timeline for migration to the five clouds, Beloussov says he sees it happening “very quickly,” being spurred on by Google rolling out more offerings and applications, and by Microsoft’s cloud presence. As far as adoption by end users, Beloussov sees (as do most observers) that it already is happening with small to medium businesses. He points out that many SMBs are running their Web sites or e-mail on a host’s servers, as well as increasing number of applications.
“What we see is tiny businesses will host their infrastructure in the cloud. Companies with 10 to hundreds of employees will host part of their infrastructure on a cloud service -- for example, e-mail or certain vertical applications -- while running other things internally. For companies with up to 500 employees, it will happen in five years,” he states. “For big organizations, it will take probably 10 years before they’re running a major part of their infrastructure in a type of cloud. They have some huge concerns. There are some security objections to moving data offsite, and often resistance from IT. Also sophistication of infrastructure and multiple SLAs makes it difficult to move to a hosted environment.”
But any predictions are up questionable with the global state of affairs, and the current financial situation might actually accelerate the timeline. “It could actually help drive businesses into the cloud because they can avoid all those capital expenses and upfront investments,” Beloussov says. “With technical budgets so tight, that will drive people toward hosted applications and services.”
That could be one small silver lining in the dark financial cloud.
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.