Our AWS Educate service primarily offers EC2 instances for faculty and students, but over time we've been able to offer more AWS services within the accounts we provision alongside EC2.
Critical Concepts
AWS services are first offered on a trial/provisional basis, until we gain experience and confidence in managing them.
- Fully supported services:
- We are very familiar with the service and offer full support for troubleshooting nearly all aspects of the service on our end.
- Trial/provisionally supported services:
- Depending on how limited resources are, support for troubleshooting may be limited due to lack of familiarity with the service.
- Due to inability to control certain aspects of a service (usually billing/permissions), usage may be throttled.
We support the ability to access and make use of Amazon's technologies, but we do not offer training on each individual service. (i.e. we won't be able to teach faculty how to use Sagemaker)
Individual Vs Shared
There are two fundamental types of AWS Educate accounts: individual and shared. We will help you decide which account structure is best for your course, but it's important to understand the differences between the two as an instructor.
Shared
- All students share the same account but have different roles.
- Students can see each other's resources.
- An example: they can see each other's S3 objects and access them.
- EC2s are an exception--they can only be started, stopped, and terminated by their creator (and the professor).
- Make up the majority of our AWS Educate accounts.
- If students are working in teams, they can share SSH keys so that they can access the same EC2 instance.
Individual
- Each student receives their own AWS account.
- Students can only see/interact with their own resources.
- If you require a service outside of EC2/S3, you will most likely need individual accounts.
Teams
The individual account structure has much better support for students working in teams. If requested, all students in a team get access to the same AWS Account with the same resources and privileges. For example (diagram attached below), Students in Group 1 can access AWS Account 1 and all of its resources, and Students in Group 2 can access AWS Account 2 and all of its resources, with each team having the ability to create, stop, terminate, etc. the resources in their account.
To see how to add your students to teams, see "Add Groups to a Course" on the Instructor Guidance for DSMLP Article. For TAs, who cannot be added to a course, we recommend using your course's grader account. If that is not possible, please contact us in your course's ServiceNow Ticket.
AWS Services
Notice: outside of EC2 and S3, these services need to be explicitly requested in your EC2 request for us to give you access to them.
Service Name
|
Works on Shared? |
Works on Individual? |
Full Support? |
Service Summary
|
Notes
|
---|---|---|---|---|---|
EC2 |
✔️ |
✔️ |
✔️ | Cloud Virtual Machines |
Specific instance types are supported. Additional instance types can be provisioned on a pilot basis after consultation. |
S3 |
✔️ |
✔️ |
✔️ | Cloud Storage | On shared instances, students can see their classmates stored files in S3. Individual may be preferred. |
Lambda |
❌ |
✔️ |
Limited | Serverless Functions | Consultation required. |
Sagemaker |
❌ |
✔️ |
✔️ | Data Analysis | Consultation required. Consider using DataHub! |
Sagemaker: Elastic Inference |
❌ |
✔️ |
Limited | Additional Computing Resources for Sagemaker | AWS Approval per each account is required. |
ECS |
❌ |
✔️ |
Limited | EC2 Manager | Consultation required. |
Fargate |
❌ |
✔️ |
Limited | Serverless Compute Engine | Consultation required. |
EKS |
❌ |
✔️ |
Limited | Cloud Kubernetes | Consultation required. |
EMR |
❌ |
✔️ |
Limited | Big data platform | Consultation required. |
Sumerian |
❌ |
❌ |
No | Create VR scenes. | Currently not allowed due to the broad permissions that Sumerian requires to run. |