Difference between revisions of "CloudStack"

From RCSWiki
Jump to navigation Jump to search
Line 20: Line 20:
 
|-
 
|-
 
|CPU intensive workloads
 
|CPU intensive workloads
|48 CPUs per node, 100's of nodes
+
|Yes; 48 CPUs per node, 100's of nodes
|1-8 CPUs per VM
+
|No; 1-8 CPUs per VM
 
|-
 
|-
 
|Memory intensive workloads
 
|Memory intensive workloads
|Up to 2TB memory per node
+
|Yes; up to 2TB memory per node
|Up to 16GB memory per VM
+
|No; up to 16GB memory per VM
 
|-
 
|-
|Storage availability
+
|High storage requirement workloads
|Shared petabyte scale storage available
+
|Yes; shared multi-petabyte storage
|<1000GB per account
+
|No; up to 1TB per account
 
|-
 
|-
 
|Data classification
 
|Data classification
Line 36: Line 36:
 
|-
 
|-
 
|Customized software requirements
 
|Customized software requirements
|Use singularity containers
+
|Yes; use singularity containers
 
|Yes
 
|Yes
 
|-
 
|-

Revision as of 20:08, 2 June 2022

Apache CloudStack is an Infrastructure as a Service (IaaS) offering that allows researchers to quickly deploy virtual machines for research projects. This service is part of Research Computing Services' Digital Research Infrastructure (DRI) and is free for all University of Calgary researchers and principal investigators.

Use cases

CloudStack allows you to create virtual machines for a wide range of workloads and use cases, including:

  • Running a internal or public facing web site
  • Running a database
  • Experiment with new software tools
  • Test out the latest release of a software package

Please note that CloudStack is offered as a research environment and is supported as such. For workloads that demand high availability and high uptime, this may not be the appropriate choice. Researchers are solely responsible for maintaining any VMs they deploy. The expectation is that the owners of the VMs will be patching and doing other maintenance work on a regular basis. We will be more than happy to provide guidance on this process.

Contact us at support@hpc.ucalgary.ca for details.

Differences between RCS HPC and CloudStack

There are some overlaps between the CloudStack offering and our existing High Performance Computing (HPC) cluster environment.

RCS HPC Cluster CloudStack
CPU intensive workloads Yes; 48 CPUs per node, 100's of nodes No; 1-8 CPUs per VM
Memory intensive workloads Yes; up to 2TB memory per node No; up to 16GB memory per VM
High storage requirement workloads Yes; shared multi-petabyte storage No; up to 1TB per account
Data classification Level 1 & 2 (ARC), Level 3 & 4 (MARC) Level 1 & 2 only
Customized software requirements Yes; use singularity containers Yes
Persistent software or services No; time limited jobs only Yes
Managed environment Yes No; self managed VMs only
Research support by analysts Yes Limited

Not sure if you need a virtual machine or a compute cluster? Is this the "Cloud"? Contact us at support@hpc.ucalgary.ca and we will assist you in using this service to support your research goals.

Getting access to CloudStack

If you are a researcher or principal investigator, please review our CloudStack End User Agreement and then request a CloudStack account through Service Now.

Once your account is ready, please refer to the CloudStack User Guide for details on using CloudStack.

Note

CloudStack is a research environment and is supported as such.

Researchers have a great degree of freedom in how they use CloudStack and they are solely responsible for maintaining any VMs they deploy. The expectation is that the owners of the VMs will be patching and doing other maintenance work on a regular basis. We will be more than happy to provide guidance on this process. Contact us at support@hpc.ucalgary.ca for details.