Mechanical engineering company automates hybrid cloud

Run workloads on-premise or in the cloud providing greater flexibility

Our customer, a globally operating manufacturer of machines for process automation and the chemical industry, is expanding rapidly. This expansion is taking place through both organic growth and acquisitions.

The CTO explains the challenge as follows: "Through the acquisitions, we constantly face the problem of integrating new control software and other systems into our infrastructure."

"Our goal was to make our development teams worldwide largely independent of the infrastructure department," explained the CTO. "We aimed to automate the deployment of development systems, so that we wouldn't have to rely on an employee being available around the clock to support our developers globally."

the challenge

"We also wanted to automate the workload placement" explains Yves Sandfort, comdivision's project manager. "The customer intended to implement a cloud strategy that provides flexible operation of workloads, either on-premises in their own data center or in the cloud. From the outset, we made it clear that not every workload is suitable for the cloud. For example, if a machine needs to be directly connected via cable, this app cannot be moved to the cloud. Additionally, one must also consider how much traffic the app generates and in which direction it goes. Inbound is cheap, outbound is expensive," Sandfort chuckles.

This led comdivision to integrate a function for the initial workload assessment into the self-service catalogue that developers should have access to. Yves Sandfort explains the customer's vision: "The developer goes to the self-service portal and gives a preliminary assessment so that it can be determined on this basis whether the workload should be hosted on-prem or off-prem."

the solution

To achieve the required flexibility, comdivision decided to integrate a hybrid cloud with VMware Cloud on AWS: “We developed a concept for the customer in which a self-service catalogue with VMware vRealize Automation is provided. Here, we can not only centrally manage project-based policies, but also cost calculation and resource usage" Sandfort explains. "With the VMware Service Broker, we can store VMC on AWS machine blueprints and also integrate AWS-native development tools and services."

Policies are to be developed to allow the initial deployment to be carried out on-premises or at VMC on AWS based on a rating of 1 to 10. "The nice thing is," says the CTO, "that we can still 'bring the workload home' afterwards, if it turns out that the traffic is higher than expected!"


"In the next step, we want to present resource planning with vRealize Operations," Sandfort reports, "so we can further improve capacity planning and transparency”. Sandfort concludes with a wink: "But one thing at a time."



Ask Yves:

* We will process your email in accordance with our Privacy Policy.
Thank you! Your message has been sent!
Oops! Something went wrong while submitting the form.