Many people who are considering taking their IT operations into the cloud find themselves overwhelmed by the new technology that they’re supposed to understand. Cloud providers rattle off terms like SaaS or PaaS like you should know what they mean, when really, it all sounds plain confusing.
With this in mind, Bernard Kur, head of product at Global Micro, has provided this handy guide to the language of the cloud, so that when you’re investing in the cloud, you know what exactly it is that you are buying “as a service”.
Infrastructure as a service (IaaS)
IaaS is perhaps the most straightforward of the cloud “aaS” (as a service) offerings. It is simply a provision model in which an organisation rents infrastructure – storage, hardware, servers and networking components – and pays for the use of the infrastructure as and when they need it. The provider is responsible for running, housing and maintaining the infrastructure.
Software as a service (SaaS)
We hear a lot about SaaS as a growing trend. If it’s not a part of an existing IT strategy, companies are thinking about it. SaaS is simply a software distribution model in which the applications are hosted by a service provider and made available to a customer over a network – typically the internet.
This negates any infrastructure requirements on the client’s side, and allows for a price-per-user payment model, which works out to be significantly more cost effective and easier to manage for the end user.
Platform as a Service (PaaS)
This model is where the service provider provisions a fully functioning computing and solutions stack on which applications are deployed. The providers are responsible for the servers, the networking and the storage, as well as managing the levels of scalability. PaaS is an operating system in which all the necessary applications are installed, rather than just providing raw technical infrastructure.
Applications developers would use this piece of the “aaS” puzzle, as with nothing to install, they could simply start coding.
A public cloud is simply the “regular” cloud, which is open to everybody on the internet.
A private cloud is the deployment of infrastructure in a provider’s data centre. The private cloud is not a shared infrastructure; it belongs only to the organisation paying for it. However, the maintenance responsibility for the infrastructure lies with the cloud provider.
Hybrid cloud is used when the customer may need both a local server running specific applications and a cloud service that hosts additional applications. In such a situation, the two are often configured for interoperability.
The two infrastructures can communicate with each other, with a layer of security in between preventing breaches of sensitive information, and the cloud provider is responsible for the maintenance of the hosted infrastructure.
Obviously the beauty of cloud, and more specifically of hybrid, is that this model can look any way it needs to, with aspects of the infrastructure existing in one environment or the other. It’s also often a good way for organisations to start testing the cloud environment without making a full commitment upfront.
“Now that you speak cloud, you are free to take to the skies, knowing that you will understand what the service providers are trying to sell you,” says Kur.
“Of course, terminology is only one aspect of understanding a cloud agreement. You should also make sure that you are buying cloud services from a reputable certified local provider with a good track record (ideally offering a financially based SLA) and good partners, who address any security concerns that you may have to your satisfaction.”