Software company

Development and Operations of a Private Cloud Platform for an International Software Company

Cloudification supported the development, scaling, and ongoing operations of a private cloud platform based on OpenStack and Kubernetes across more than 15 production and QA regions.

100,000+

Employees worldwide

15+

Production & QA Regions

Unternehmen

Germany · ERP & Enterprise Software · 100,000+ Employees

The client is a large international company from Germany that develops ERP systems and enterprise software. The company employs more than 100,000 people worldwide.

The company uses a private cloud platform based on OpenStack and Kubernetes. The infrastructure spreads across 15 production and QA regions.

Challenge

Cloudification was involved in the development, scaling, and operation of the client’s cloud platform.

The scope of work included extending OpenStack functionality, developing new drivers and integrations, upgrading OpenStack releases, and modifying the platform architecture.

Day-2 operations were another important part of the project, covering the work required after the initial deployment of the infrastructure. In a large private cloud environment, the platform continues to evolve after launch: new component versions are introduced, architectural changes need to be implemented, and the infrastructure must be kept up to date and stable.

In this project, these activities covered more than 15 production and QA regions.

KEY CHALLENGE

Continuous platform development and day-2 operations across 15+ production and QA regions

Project specifics

One of the main characteristics of the project was the scale and complex upgrade and release processes.

OpenStack consists of multiple interconnected services responsible for different functions within the cloud platform. As such an environment evolves, changes can affect not only an individual component but also related integrations, configurations, and infrastructure management processes. Due to scale and custom changes made in the source code of OpenStack, the upgrades between releases often required porting of hundreds of patches between the branches, complex end-to-end testing and build of new images with custom dependencies.

The presence of multiple production and QA regions adds another layer of complexity. The largest regions are typically the last to receive updates and the ones where things can go wrong due to scale. The smallest regions run hundreds and thousands VMs, whereas the largest production regions run tens of thousands of VMs. The configurations, number of services and replicas thus had to be tweaked for larger regions to sustain the load.

In this project, the work involved not only operating the existing environment but also continuously working on the OpenStack platform itself, from developing new capabilities to upgrading releases, modifying the architecture and deploying new services.

Solution

The Cloudification team contributed to the development of the client’s OpenStack platform at several levels.

As part of the project, new OpenStack features, drivers, and integrations were developed and contributed upstream. This made it possible to adapt the platform to the requirements of the existing infrastructure and extend its functionality.

The team also performed OpenStack upgrades from Train release onwards. For a large distributed infrastructure, upgrading a cloud platform is a separate technical task: any changes have to be verified in the context of the existing architecture, custom developments and integrations with other components.

Another area of work involved changes to the platform architecture and the implementation of complex modifications. Unlike a local change to a single component, such tasks require consideration of the relationships between different services in the OpenStack ecosystem and the subsequent operation of the updated infrastructure.

After implementing changes, the team continued working with the platform as part of day-2 operations and development support. Cloudification therefore contributed not only to the development of individual solutions, but also to the ongoing operation and evolution of the infrastructure across more than 15 production regions over the years.

Result

Cloudification supported the development, scaling, and ongoing operations of the client’s private cloud platform based on OpenStack and Kubernetes.

The project included the implementation of new OpenStack features, drivers, and integrations, OpenStack release upgrades, and changes to the platform architecture.

The implementation of complex changes and day-2 operations covered more than 15 production and QA regions.

The project enabled the continued development of a large distributed OpenStack infrastructure, covering not only its operation but also the evolution of the cloud platform itself.