Research Computing Centre

Cloud Computing

  1. Home
  2. Systems

At RCC we have built a team of experts in cloud computing and are working with our partners (QCIF and ARDC) to provide simple and effective solutions in research.

What is cloud computing?

Cloud computing is a term that can be applied to several different approaches to delivering computational resources. The main characteristic of cloud computing is that you are sharing “someone else’s computer.” This means that rather than having direct access to the physical infrastructure, you consume an abstraction of that infrastructure instead. Many instances of these abstractions may co-exist on the same physical infrastructure and not all of them will be owned by a single person. Managing these instances is achieved by either a web interface or via an application programming interface (API). Often there are command line tools or libraries available to make it easier to use the API.

There are three main classes of cloud computing:

  1. Infrastructure as a Service (IaaS): In this case the abstraction is done of infrastructure; a virtual machine, storage, and/or networking. Some of the large public cloud providers are Amazon Cloud Services, Microsoft Azure, Google Compute Engine, Oracle Cloud, and IBM cloud services. The primary open-source solution for IaaS is OpenStack.
  2. Platform as a Service (PaaS): In this case, infrastructure has been abstracted away to allow you to focus on creating a service. Examples of this are web hosting, database hosting, virtual desktop services and Kubernetes.
  3. Software as a Service (SaaS): In this case the abstraction is an application. Examples of this are Office 365, JupyterHub notebooks and Workday.

Note that some cloud services do not neatly fit into these categories and may provide one or more of these offerings.

Due to the on-demand nature of cloud computing, most company pricing models for cloud computing are based on per second and per resource billing. However, pricing can become complicated as quite often several services are required to deliver a solution to a problem.

Also, there can be different options that can have a significant impact on what you pay. Predicting cloud costs can be a challenge for even the most careful of users.

Cloud computing embodies several concepts:

On-demand and scalable access

On-demand and scalable access means users should be able to scale resources dynamically without the need to purchase hardware. You may need one compute core today, but 1,000 tomorrow.  This is called “elastic computing.”

Application services

Application services mean the Cloud provider might expose high-level services that meet a business or research need. This currently works well for tools like email and calendars, but the number of applications of value to researchers is growing all the time; some of these will be very research domain specific.

Location transparency

Location transparency means just that —knowing where your services are being sourced. There are applications where physical location is important. For example,  you might want data to be held in a particular data centre. Thus, most Cloud providers allow you to indicate a locale that restricts the actual domain.

Virtual machines

Virtual machines are a technical trick that simplify the deployment of new services and are part of providing rapid scale-up and scale-down. Putting it simply, it means you can build a software stack locally and deploy it on as many physical machines as required.

Commercial Clouds also include pay-as-you-go economic models. This means you pay only for the infrastructure you use — whether that’s machine time or data storage space. This is useful for business applications, but not as commonly used in research circles yet and is highly dependent upon the use case, scale and context of the research domain, applications and desired end goal.
 

ARDC offers eResearch infrastructure in three key areas:

  • Research Cloud
  • Virtual Laboratories
  • eResearch Tools.

The ARDC Nectar Research Cloud is a computing resource for all Australian researchers. It offers virtual machines where researchers can develop and deploy applications and collaborate in a uniform environment with controlled sharing of data.

At UQ

UQ researchers have access to data storage on the Queensland node of the national ARDC Nectar Research Cloud called QRIScloud, operated by QCIF and RCC.

QRIScloud

QRIScloud is a large-scale cloud computing and data storage service. It aims to stimulate and accelerate the use of cloud computing by researchers in all disciplines, and enhance collaboration with other researchers nationally and internationally.

Using QRIScloud, researchers can compute using single instances or large virtual clusters. They can also host web services, store research dataset (even very large datasets) and share existing data across research communities. 

QRIScloud allows researchers to:

  • share dataset with collaborators worldwide
  • manage access to their data collections
  • leverage data collections stored in state and national nodes
  • integrate access to Queensland-based High-Performance Computing facilities and specialised cloud services
  • access virtual labs with national communities
  • launch on-demand compute services
  • host web services
  • access and use a wide range of existing services, tools and applications.

For more information on how to apply for data storage or compute resources on QRIScloud, go to QCIF's QRIScloud website.