November 16, 2021
Just as a picture can speak a thousand words, a server image can hold terabytes (or more) of data. These images can be used as a backup or as a template for replicating the setup. By creating a “golden image,” or template that is compliant, consistent and easily replicated, cloud resources can be set up more efficiently – allowing more time for innovation.
With the creation of a master template or golden image as the goal, the Division of Information Technology created a Continuous Improvement/Continuous Development (CI/CD) pipeline inside its Aggie Innovation Platform (AIP) – which can be likened to the software version of an assembly line. The nine-month pipeline project was launched to create autonomous, virtualized computational server images that can be run on a diverse set of cloud environments. The resulting AIP-managed images are ready for customers to use within their linked Amazon Web Services (AWS) accounts, giving them the template needed.
Andrew Marcontell, director of the Division of Information Technology Enterprise Platforms, Cloud and Shared Services, said the project will aid system administrators in the utilization of cloud environments.
“Previously, you had to build an image directly from the hardware or an abstracted version of the hardware, which could take one to two days and had high error rates,” he said. “Now we start from a base image, integrate additional software, settings and security controls, validate it and then release it for use. This is an example of Infrastructure as Code, or IaC. Customers then take the resulting software package and can create an image in minutes with low error rates and a higher level of compliance.”
The images currently available are Amazon Machine Images (AMIs) for server-type operating systems such as Ubuntu and CentOS. Future development will include RHEL (Red Hat) and Windows. Customers can deploy the images in the AWS Graphical User Interface (GUI) and Command-line Interface (CLD), as well as through the Texas A&M Service Catalog. AIP actively encourages community involvement in the pipeline to crowd-source future functionality and aid in the release of new versions of the code.
“This project has resulted in a number of benefits,” Marcontell said. “They are so significant that we are truly at the precipice of a new era in IT system administration.”