Case Study

Cloud migration without downtime thanks to a hybrid architecture

Bayer Crop Science combines the advantages of cloud and on-premises, enabling smooth scaling of its tool QuickScan despite dependencies on legacy systems.

#Migration
#Cloud computing
#Agricultural sector

Picture

The challenge

Migrating a tool to the cloud despite on-premises dependencies - and remaining fully operational during the process.

Picture

The scenario 

Automated evaluation of ingredients

As part of the Bayer Group, which operates in the fields of health and nutrition, Bayer Crop Science develops and markets crop protection products. Very different rules apply worldwide to the safety evaluation of these. In low and middle-income countries (LMIC) that do not have sufficient capacity to fully evaluate product safety, Bayer raises the safety bar by following internal guidelines on top of local regulations, while considering the unique needs of growers in those markets.

With Quickscan, Bayer Crop Science avoids time-consuming manual risk assessment. The tool uses various algorithms to automatically evaluate the operators, environmental, and dietary safety of a crop protection formulation. A viable practical solution that caused demand to skyrocket. The on-premises solution and the outdated technology stack behind QuickScan could no longer cope with this growing demand.

The requirements

System adjustments during migration

Bayer Crop Science needed a solution that could comfortably handle the increased amount of checks without long waiting times. A cloud migration would meet this requirement . However, since various on-prem and legacy systems form the basis for the algorithms behind QuickScan, the tool could not be completely detached from the original environment.

At the same time, a solution was needed in which the tool would be available without downtime and would allow adjustments to be made easily during the migration. This is because product safety requirements are steadily evolving.

The solution 

On-premises and cloud combined

Today, QuickScan is based on a hybrid architecture. While the database continues to reside on Bayer Crop Science's on-premises server, the processing of the data in order to generate product evaluation has been moved step by step to the AWS cloud. To make this possible, experts from Comsysto Reply and Leadvise Reply containerized the individual elements of the QuickScan application using Docker. This made them lightweight and portable. They were then deployed in the cloud using Amazon Elastic Container Service (ECS). To enable the algorithms within to access the on-premises server database, the server was reconfigured along with the relevant APIs. The special firewall settings ensure that this access is secure.

Picture

Thanks to the use of GitHub, new features and customizations can now be quickly and easily implemented and deployed in an automated manner.

Picture

Thanks to the hybrid architecture, Bayer Crop Science could continue to use the tool even during the migration.

At any time, it was possible to flexibly adapt the system to new legal regulations. Today, Bayer Crop Science benefits from a future-proof, state-of-the-art solution that can handle 10 times more requests and easily scale up as needed.

Picture

Bayer is a global company with core competencies in the life science fields of health care and nutrition. Through its products and services, the company aims to benefit people and protect the environment by helping to solve fundamental challenges of an ever-growing and aging global population. Bayer is committed to making a significant contribution to sustainable development through its businesses. At the same time, the Group aims to increase its earning power and create value through innovation and growth. The Bayer brand stands for trust, reliability, and quality worldwide. In fiscal 2022, the Group generated sales of €50.7 billion with around 101,000 employees. Research and development spending, adjusted for special items, amounted to €6.2 billion.