Reduce the time it takes to obtain a virtual machine from a week to 15 minutes

HPC
Written by Wilfried Kirschenmann, on 21 August 2019

This year, ANEO is renewing its participation in DevOps REX! On this occasion, we share our feedback on concrete DevOps cases.

We met the IT department of a large French bank that identified the need to improve its time-to-market and make its developers more autonomous in infrastructure management. A significant performance bottleneck was the slow manual creation of IT components.

The request process was manual, passing through a ticketing software generating defined requests but requiring various human actions, significantly delaying the delivery of these components. The tools needed to be redesigned with a DevOps approach so that clients could create all types of infrastructure automatically, without manual intervention.

The Business thus approached the IT department to create an API interface of Infrastructure as Code type to manage the infrastructure like code and automate the construction of IT architecture without human intervention.

This major project was launched with the support of several consulting firms, including ANEO, starting with upskilling the IT department, gathering client/business needs, followed by a comprehensive study of human actions from the lowest to the highest level before implementing the interface designed to strengthen the infrastructure.

The time to obtain a virtual machine decreased from one week to 15 minutes!

But that's not the only advantage of such a project... Let us tell you the others!

We were involved in this creation and developed numerous features gradually reducing the time to obtain a virtual machine (and other infrastructure components) to about fifteen minutes, compared to sometimes a week with the previous system.

In addition to making life easier for developers and allowing them to be completely autonomous in designing and implementing their infrastructures, the IT department was able to reduce costs by dedicating fewer people to these tasks; reduce its time-to-market; ensure production deployments are more reliable, thus increasing its profits; position itself as a technological leader, also attracting IT talents who might have otherwise opted for more modern IT startups.