- Defining cloud ...
Defining cloud and software deployment models
By: Bryan Hill
The term 'cloud' is often used loosely to denote different services and infrastructures. As a key enabler of cloud services we see a lot of confusion in the usage of terms around cloud and this can be problematic when one person’s definition of IaaS or hybrid is different from another’s. For clarity, we've defined the main cloud models that are relevant to digital video production and distribution.
Virtualisation is the creation of a virtual version of a device or resource, such as a server, storage device, network or operating system where resource is divided into multiple execution environments. Virtualisation is most commonly associated with servers — the partitioning of a physical server into smaller virtual servers — a technique that helps organisations to make more effective use of compute capacity by running multiple applications on a single physical server.
Infrastructure as a Service (IaaS)
This is the basic cloud service, providing resources such as virtual machines (VMs), network load balancers, storage and connectivity. In this model, computing resources such as CPU, memory, storage and network are added to a VM, on which an application is then deployed. The underlying infrastructure is managed by the cloud (or IaaS) provider; and the customer of the infrastructure maintains their own applications on it.
Platform as a Service (PaaS)
PaaS is the next layer up from IaaS. With PaaS, computing platforms, operating systems, databases and applications are delivered by the cloud provider. Customers deploy onto, and developers write code for, the platform. This is often a highly automated and scalable environment.
Software as a Service (SaaS)
The top layer is where the application software is installed in the cloud and accessed by the customer. The SaaS provider manages the software and all the underlying technology, effectively renting the software to its customers. IaaS, PaaS and SaaS can all be deployed in public, private and hybrid cloud environments.
A public cloud infrastructure is one that is made generally available to enterprises and individuals by its owner. Public clouds include those owned by Amazon, Microsoft, Google and IBM.
Pros: initial cost savings, flexibility, scalability speed and ease of deployment.
Cons: possible lack of control, perceived security concerns, possibility of higher costs over the long term compared with dedicated infrastructure.
A cloud infrastructure operated for an individual organisation that uses it to run its own applications, to deploy software, or to offer SaaS to other enterprises. A private cloud can be installed on the organisation's own premises or in a third-party data centre.
Pros: the benefits of public cloud without any concerns about control and transparency; can be designed to meet specific requirements.
Cons: Less flexibility, less scalability and higher initial costs than public cloud.
A combination of two or more compute platforms (for example, in-house infrastructure and public cloud) that are separate but connected, and enable data and application portability and interoperability, for example, bursting onto the public cloud to handle peaks in demand that exceed trended norms.
Pros: the best of both worlds — flexibility, scalability and savings with security and control.
Cons: the most complex solution to set up and manage, requiring specialist expertise and excellent connectivity.