Cloud Computing Services
Red Cloud is a subscription-based Infrastructure as a Service cloud that provides
root access to virtual servers and storage on-demand. The cost is less than comparable public cloud instances.
- Subscriptions cover 1 core-year of virtual server usage. Persistent disk storage (volumes) backed by Ceph storage are available with more than 1.5 petabytes of raw capacity. Your first subscription includes 50GB storage. Additional storage may be purchased separately. Your subscription is only charged when you request virtual servers. You can look up your balance at any time. Our subscription model helps users avoid cost overruns.
- Unlike AWS and other public cloud CPU cores and RAM, Red Cloud CPU cores and RAM are not oversubscribed or shared with other users. Users have full access to the underlying physical processor and memory without competition, resulting in fast and consistent performance.
- Users can request instances with up to 28 cores and 240GB RAM. Each instance is a virtual machine (VM) that deploys in seconds in the cloud. A variety of VM configurations are available.
- NVIDIA T4 and V100 GPU instances are available.
- Users control access to their instances by port numbers and IP addresses for maximum flexibility and security.
- We use the OpenStack platform. Users can manage their cloud services using a Web console, command line clients, or any development library supporting the OpenStack API.
- Linux and Windows Server operating systems are available.
- There is no charge for network traffic.
- MATLAB Parallel Server is available on Red Cloud. Users must be familiar with MATLAB Parallel Computing Toolbox (PCT) to use MATLAB Parallel Server. PCT is included in Cornell's site license. A PCT tutorial is available or you may add our consulting services to your project when you start it up and we will help you with PCT or any other problems. In addition to PCT, specific MATLAB clients must be installed on your local workstation. See assumptions for the MATLAB clients we support and other project start-up information.
- Note: Red Cloud is not suitable for Cornell high-risk or restricted data, or for state or federally regulated data. Data stored on Red Cloud is not backed up by default; users are responsible for their own backups. EZ-Backup is a solution available for Cornell users.
- 1 Red Cloud subscription = 8,585 core hours
- 8,585 core hours = approximately 1 core running for 1 year
- Core hours are like a Starbucks card. You can use a subscription up fast by using many cores over a short period of time or slowly by using, for example, just 1 core for 1 year. They are good until they’re gone. The subscription model prevents billing surprises.
Subscriptions and Exploratory Accounts for Cornell Users
- Cornell students, faculty, and staff are eligible for a free 4-month Red Cloud exploratory account
(165 core hours, 50GB storage, 1 hour consulting).
- Cornell subscriptions are $300.00 (8,585 core hours and 50GB storage with first subscription).
Additional storage is $100.00/TB/year or $0.10/GB/year. No network charges. Discounts are available for multiple subscriptions (8 subscriptions = 10% discount; 16 subscriptions = 20% discount; 32 subscriptions = 30% discount; and, 64 subscriptions = 40% discount).
- Optional consulting, e.g., building your virtual machine image for you so it’s all set and ready to go, is
help for assistance in estimating consulting hours.
- To start your subscription, complete a project request form. A Cornell University account number is required to charge back to.
Subscriptions for External Organizations
- Red Cloud subscriptions are available to
external organizations (start-ups, business and industry, universities, R&D centers, not-for-profits, etc.) through fee-based CAC Partner Program memberships.
- Optional consulting hours may be included in your Partner Program membership to build a cloud image for you, containerize your application for portability, support your web-based application, architect a private cloud, etc.
- Contact Paul Redfern, Partner Program Director.
- Visit the Red Cloud wiki.
- Still have a question or two? Contact us. We’ll be happy to help.
- In-depth consulting is available for a fee.