Cloud computing
Our cloud computing environments are Infrastructure-as-a-Service clouds, meaning that users are given access to computing resources that allow you to launch and configure virtual machines (VMs) at any time. Cloud computing resources:
- are a flexible solution for more customized project needs
- let you set up a virtual server that can be used for whatever you want, whenever you want
- require technical skill to set up and maintain
Cloud computing resources are managed by the Digital Research Alliance of Canada ("the alliance"). Our alliance resource page will help you create an alliance account, understand user roles, and create cloud VMs. Your cloud project will be set up on a cloud like Arbutus, a dedicated research cloud hosted at UVic.
Details
- Type
- HPC service
- For
- Principle investigators (PIs) and sponsored users
- Cost
- Free
Use cases
Cloud resources can be utilized in multiple ways. Many researchers are using our clouds to host:
- Django and Flask websites
- Virtuoso servers
- topic modeling and text mining explorers
- custom configured Jupyterhub environments
Other researchers are using cloud resources to run browser-based research tools like:
- data portals
- monitoring platforms (like health applications)
- custom platforms for organizing their HPC cluster batch jobs and post-processed data
Examples
- The Mountain Legacy project uses cloud computing to capture change in Canada's mountains.
- The ATLAS Tier 2 site at UVic uses a Kubernetes cluster to process computing tasks for particle physics research.
Cloud computing storage
Cloud computing storage is available for researchers using Alliance cloud computing instances. This type of storage makes your data accessible to the cloud virtual machine resources you've set up for your research.
- Can store up to 30TB per research group in a default allocation.
- Data is automatically backed up nightly.
- Service is located in secure data centres at UVic.
- Used for research data used with a cloud computing project.
- Requires a system administrator to set-up and maintain.
- Multiple types of storage to choose from.
Alliance resources
Cloud computing resources are free for all Canadian researchers with a cloud project. You'll be assigned your default allocation of resources when your cloud project is approved. These resources include:
- (virtual) CPUs/GPUs,IP addresses, and RAM to create different virtual machines (VMs).
- 10TB storage quota.
You can create as many VMs as your resource allocation allows. If your project needs additional resources, you can submit an application to the resource allocation competition (RAC). We put together some resources to help you with your RAC application.
Setting up cloud computing requires a system administrator with Linux and Windows knowledge to set up and maintain. To get started, you'll need to:
Read more about cloud computing support.Commercial cloud
If you're looking for alternatives to the Alliance cloud computing resources, there are some commercial cloud computing options available as well.
We strongly recommend planning a consultation with us before choosing a commercial cloud platform. There may be more sustainable options available for you to use for free.
Limitations of commercial cloud
- Our UVic team can't provide any technical support.
- Costs are difficult to predict and change frequently, so you may run out of free credits quickly.
- Terms of use may change with little warning.
- Can't guarantee data storage in Canada.
Here are some of the commercial cloud platforms offering free or subsidized resources for academic researchers in Canada:
- Google provides cloud credits for PhD, postdoctoral researchers and faculty. Users can apply for access.
- AWS Public Sector Cloud provides cloud credits for researchers as part of its Credit for Research Program.
- Oracle provides cloud credits for its research cloud.
Get help now
For help with research cloud computing, you can:
- visit the Alliance tech support wiki page
- email the Alliance cloud support team
- contact Research Computing Services to consult with a research computing expert