The scramble for cloud standards and control

As cloud computing matures, issues such as common standards and interoperability remain obstacles, even as customers are encouraged by the cloud's commitment to better infrastructure management and cost-cutting.

In one-on-one interviews, our co-founder and chief technology officer Randy Bias, cloudscaling of the open source cloud infrastructure provider, explores the lack of cloud standards among providers and how today's cloud computing service providers can remedy them, Such manufacturers will eventually give way to a more unified cloud world.

TechTarget: Many major cloud provider application interfaces are incompatible with others ' APIs. Can this change?

Randy Bias: What I see is that the infrastructure that is critical to the company will eventually be built into standards, APIs, and consistency. Without these, it is difficult to create a large number of standards and stacks over the upper API.

TechTarget: Which cloud maker is likely to win the lead in this battle?

Randy Bias: You can complain about whether Amazon Web services or VMware are the right models, but we can say with certainty that they dominate the market in their respective parts, leading to a certain amount of consumer awareness.

Some people, however, feel uncomfortable with the dominant position ... They must justify other standards, platforms and APIs. But 90% of the public cloud storage capacity of Amazon and 90% of the Enterprise virtualization cloud belongs to VMware, in this case, I can hardly believe that there will be other standards, which is meaningless.

TechTarget: Is it a problem that the APIs of the public cloud and private cloud providers are incompatible with one another?

Randy Bias: There are two different architectural approaches. The VMware model is what we call an "enterprise virtualization Cloud" that enables the acquisition of existing islands of enterprise-small islands of all different hardware, software, networks, and storage architectures-to rebuild and virtualize them. But in this way, economies of scale cannot be achieved.

This approach is therefore driven by greater server consolidation and greater usage requirements.

Amazon's architectural approach is a better reminder of what the Internet giant has been doing. Instead of building an island, they build layers of cakes, a whole stack of services that support each other, from hardware to software. Then, at the software layer, there is a whole stack of applications that can take advantage of the entire heap. We call this the "resilient infrastructure cloud." "It can meet the greater scalability of applications and infrastructures, rather than managing a large number of islands, but it enables applications to reflect the underlying infrastructure."

TechTarget: How do these cloud solutions actually differ?

Randy Bias: Different types of applications and workloads for each application are very different.

VMware customers have database servers like Oracle, and the servers are old and have no basic replication capabilities. They rely on the infrastructure for a great deal of practicality, and the design of applications proves that these infrastructures never regress.

For customers like the Amazon (Netflix), we see a completely different experience. They are redesigning their apps and they have made the basic assumption that the server will sooner or later be machine, just a matter of time. So the question is how do you design your application to handle these issues?

People's perceptions have changed dramatically; people now think it's important to extend apps horizontally like Facebook's big platform.

TechTarget: Will the company revolve around a common standard?

Randy Bias: There's a certain amount of concentration now, but frankly, customers are everywhere. It is too early to understand that there is a disconnect between the two distinct patterns. Consumers believe that when they create a VMware environment that allows them to start virtual machines, they are also rebuilding the Amazon environment--apart from the inability to provide scale or flexibility. They found that it was not good enough for application developers to succeed internally and continue to go to Amazon. People will continue to fail and learn how to face difficulties through failure.

What we really lack is a group of solution providers who can also wrap and deliver Amazon equivalents, not just APIs.

TechTarget: If so, what role does open source play in cloud standards?

Randy Bias: Because of this particular disconnect, open source code and open standards are in the lead. If you go to a lot of really big companies like Google and Facebook, you'll find that all of their systems are open source. That's because of cost economics, but in many ways it's also a matter of controlling your own destiny.

At the same time, you will see strong resistance from companies that rely on software and hardware vendors to make a profit, and you will feel uncomfortable being stuck between these suppliers. Companies are also beginning to notice that open source is the level of maturity they can accept.

TechTarget: How is the cloud different from its predecessor?

Randy Bias: This almost destroys the fact that the client-server and enterprise computing paradigm replaces the mainframe paradigm.

We've grown from a huge, single, one-sign-on iron box to running one or a few applications to more distributed--a shared responsibility between IT staff running centralized servers and users running laptops or desktops--to a more fully distributed model.

(Responsible editor: Schpeppen)