Cloud computing models: Public vs. private vs. hybridDate: Nov 15, 2010
There has been a good deal of debate in the cloud computing marketplace over the three major cloud models and the characteristics they possess. All three examples have their own place in the cloud world, and the benefits each offer can appeal to different enterprises in different situations. And when you include Infrastructure as a Service (IaaS), Platform as a Service (PaaS) and Software as a Service (SaaS) in the equation, it's enough to have IT managers everywhere ripping their hair out.
So when you and your enterprise decide to pursue cloud services as a legitimate computing option, make sure you're properly informed on what the cloud has to offer. In this webcast with executive editor Jo Maitland, find out what makes up the three major cloud models: public cloud, private cloud and hybrid cloud, and the differences between public vs. private clouds.
Read the full transcript from this video below:
Cloud computing models: Public vs. private vs. hybrid
Jo Maitland: Hello. Welcome to our webcast on Cloud Computing Models: Public vs. Private vs. Hybrid. This is the first lesson in our interactive classroom on Cloud computing. My name is Jo Maitland. I am Executive Editor of the site SearchCloudComputing.com, at TechTarget. You can reach me at email@example.com.
To get into the definitions and to be clear about what we are talking about, there is a lot of debate over these different areas of the Cloud market. Here is a diagram that more clearly explains the different infrastructure alternatives.
Public Cloud is probably the one that most people are familiar with. This refers to external resources, compute resources, server storage networking that you can actually buy from a service provider. The most well-known service provider in the public Cloud computing space is Amazon and the Amazon Web Services business. Of their services EC2 is the most well-know and well-used service, which is their servers on demand, which is just buying servers as you need them. This works in a way where those resources are dynamically provisioned, and you buy just as much as you need. You buy an instance, or a virtual machine, you start it up and you stop it when you are done running that job. Other companies in this space are Google. Google has something called Google App Engine and also Microsoft has something called Microsoft Azure. The difference with these last two, Google App Engine and the Microsoft Azure service, these are actually platforms-as-a-service. In this instance, as the user, you do not care about the underlying storage or server resources. You actually just deal with the application tools above that to build your app. In the case of the Google App Engine, you can use Python, and you can also use different versions of Java. You write your application, and then the resources underneath scale according to what your application requires or what your web service requires. You do not actually have to worry about the hands-on scaling of that infrastructure; that is taken care of for you, and the same with Microsoft Azure, though in the case of Azure it is the .NET tool sandbox that you are playing with to build your application. That is how the public Cloud market breaks up between infrastructure-as-a-service and platform-as-a-service.
In between here is something that is getting a lot of press these days called Hybrid Cloud, and Hybrid Cloud consists of internal and external resources. The most well-known example here today is VMware building out private Clouds, so people that run VMware internally at their organization can bridge to an external service provider running VMware services on their infrastructure, so you can securely connect your internal VMware environment to a public external VMware environment. Some examples of those are Verizon, which offers VMware servers on demand as does Terremark, as does Savvis. There are a few companies, services providers coming on board offering that service now.
It is still very new and early and not many people are doing this, it is hard to find anyone today, but the expectation in the long run is that this will be typical for most enterprises. There will be certain applications that you will keep in-house and run on your private infrastructure, and then many other applications that you will be perfectly happy running in a public service provider like Verizon or Terremark. There is a lot of talk, a lot of debate around that taking off right now.
More controversially, in some ways, is this notion of a private Cloud which is an internal Cloud. It is actually extraordinarily hard to find anyone today that has a true private Cloud, in the sense of an Amazon Web Services-like Cloud, internally, where the entire thing is automated; the entire infrastructure is virtualized and automated, and the services are delivered through a self-service portal that is charged back and all those other good things that public Clouds enable. There are probably a handful of companies around the world today that can actually put their hands up and say they are doing that. Most of those are actually banks, and they have actually rolled their own software and built these things entirely from scratch with off-the-shelf components and in many cases, open source software.
The big deal amongst the IT vendors is that private Clouds deliver a lot of the benefits of public Cloud computing without some of the pitfalls, so they capitalize on data security, corporate governance and reliability fears and concerns that Enterprise IT organizations have about the public Cloud, and that is very justified. Private Clouds have been criticized on the basis that users still have to buy, build, and manage the Cloud themselves, and as such, they do not benefit from the low upfront capital costs and less hands-on management. They are lacking the economic model that Cloud computing has made such an intriguing concept to organizations today.
Those are the three key different types of Clouds that you can build, with most of the action happening, most of the talk is happening around this hybrid idea. A lot talk around private Clouds, but not very many actually in existence. Then lots of people testing the public Cloud for non-critical things like web storefronts, test and dev applications, and that kind of thing.
For more resources on this topic, certainly check out SearchCloudComputing.com, which covers everything in this marketplace, and then our two virtualization sites: SearchVMware, which has everything you need to know about what WMware is doing around Cloud computing and our SearchServerVirtualization site, which takes a broader look at the ecosystem around this marketplace. It also looks at some of the implications for the network, storage, and all those kinds of things, once you move into a fully virtualized environment. External resources. Outside of TechTarget, certainly check out the CloudSecurityAlliance. These guys are at the forefront of what is happening to improve security standards in the Cloud. CloudAudit is also trying to get the industry and the IT vendors on the same page in terms of standardizing how you move workloads between different Clouds and standardizing APIs, things like that.
The last two companies on this list, CloudHarmony and CloudSleuth, are two small organizations just started recently that are tracking public Cloud providers' up-time, so you will see when there are outages, you will see how long those outages are. You can compare Amazon Web Services vs. Google vs. Rackspace. There are a whole bunch of them on there, and also, now increasingly the teleco's are being added to that list, so you can see the up-time they have, their response times to outages, that sort of thing.
This was our lesson on public vs. private vs. hybrid Clouds. Thank you very much.