In my last couple of posts I’ve talked about High Performance Computing in the cloud. Specifically about extending your compute clusters to take advantage of additional compute resources when needed without having to invest in all the hardware, software, etc to take care of short term demand for additional capacity.
So, if this is something you need, how do you get started? The first thing is to determine if it's even feasible for your organization. Will your security protocols allow for externally hosted servers? Will extending to AWS or Azure meet your organization's policy guidelines for the cloud. Assuming this is ok, and that you have an account with Amazon or Microsoft where you can access these cloud resources, we can proceed. If not, you’ll need to get that sorted out first.
The good news is if the above is in place, extending your Bright cluster to the cloud is a fairly straightforward process, but there are potential pitfalls that you will want to be aware of and things that you will need to do to ensure everything works properly. To that end, we’ve created a Getting Started Guide to help you through this process. This document is broken up into four sections:
- Prepare for Deployment
- Deploy the Bright Cluster Extension
- Provision the Cloud Director
- Provision Cloud Compute Nodes
Most all of the necessary steps for deployment are done through the cloud wizard in your Bright View administrator console. From there you can enter your account credentials for your cloud provider and configure all the necessary parameters for your deployment. The Getting Started Guide will walk you through the process and help you determine the best way to set these parameters based on your specific requirements.
That’s all there is to it. If you need that additional capacity, even for just a short period of time, it’s worth checking it out.