University IT Infrastructure with NetOrca
Universities are increasingly relying on IT infrastructure to meet the diverse needs of their faculty and students. As both the number of Cloud-based enterprise solutions and the users increase, there is a compelling need for automation and orchestration, and a user-friendly interface accessible to all members of the university community to ensure efficient and effective utilization of resources.
Imagine a scenario where a professor needs to provide virtual machines (VMs) for their students. These VMs might need to be equipped with powerful GPUs for training AI models, alongside shared and consistent storage, and private GitLab repositories. Traditionally, the process of acquiring these resources involves sending multiple emails to the university’s tech department, doing some paperwork, and waiting for manual provisioning. Moreover, professors often require students to install software necessary for their coursework. The installation process can be daunting and time-consuming. In some cases, they might spend a few sessions just mentoring students through these installations, even though the installation process itself is not part of the core learning material. The IT staff, who are responsible for providing these resources for the university, may have automated certain deployment processes, there is still a need for a user-friendly system that facilitates and orchestrates resource requests within the university community. This is where NetOrca comes in.
NetOrca can deliver all infrastructure resources, both automated and non-automated, as a "Service" to users. Students, professors, and staff can "consume" these services for specified periods, such as a semester.
For instance, a professor can set up a team for a course, submit a detailed request, in a declarative format, for the needed resources, and the infrastructure department, acting as the service owner, can review, approve, and process this request automatically, respond all necessary information, such as VMs' credentials.
Moreover, to standardize user requests and ensure submissions adhere to predefined criteria, service owners can define JSON schemas for each service. This approach accelerates the entire process and minimizes conflicts and unnecessary communications with the IT department.
An Example
Consider a scenario where a Computer Vision course is being offered this semester. The professor wants each student to have a virtual machine (VM) preinstalled with relevant modules and software such as OpenCV, and equipped with GPUs to facilitate the training of deep neural networks for their projects.
The IT department provides a VM service and defines a schema specifying the required parameters: number of CPU cores, RAM, storage, type of GPU, and selection of necessary modules. The professor uses NetOrca to create a team for the course, adding TAs and students as members. He then generates declarative submissions for the required number of VMs, adhering to the service schema, and submits the request.
Subsequently, the IT staff receives the declaration with the specified requirements, processes each service item, and provides all necessary information for students to access their VMs in an automated process.
At the end of the semester when the course has finished, the consumer team submit again to NetOrca to delete all items and free the resources
Other relevant features
- The IT department, i.e. The Service Owners, can charge the students and professors for each service, and provide a comprehensive cost report, this could also work with a credit system.
- For high scale of submissions and big declarations and more supervision and approval, NetOrca can be integrated with version control softwares like GitLab and GitHub. This would also help introduce Git and declarative concepts to the students.
- For faster processing, NetOrca provides Webhooks to trigger the automation scripts directly.
Value At a Glance
Service based automation would provide:
- Large cost savings on expensive infrastructure
- Clear demarcation of customer requirements vs the deployed infrastructure.
- Statistics and cost of usage per user.
- Flexibility for students to choose their infrastructure and do this in an efficient way.
- Ability to be applied for any limited resource across any university course.