The shoe retailer has moved its databases to the Oracle cloud. The goals: higher performance, less overloaded systems and more room for innovation. […]
Deichmann has modernized its data management. With the help of a new database infrastructure based on the Oracle Exadata Cloud Service, the largest European shoe retailer wants to accelerate and simplify processes in IT operations and development.
“In the past, we procured the hardware of our databases from other providers and operated them ourselves,” says Ralf Woltering, Head of Business Intelligence at Deichmann SE. With the move to the Oracle cloud, the company wants to be more flexible with large workflows and save time and resources. The Group operates the databases for all its applications, with the exception of the North America business, with solutions from the Californian IT provider.
As hardware cycles were coming to an end and data volumes were increasing, Deichmann was looking for an alternative to its on-premises appliances. With the old infrastructure, the nightly ETL batch processes took a lot of time. ETL stands for “Extract, Transform, Load”, whereby data from different sources are consolidated in a target database. Many automated user reports from the retail business also put a heavy strain on the systems, which also had an impact on day-to-day operations. In April 2021, after a proof of concept, the retailer decided on a cloud database of around 28 terabytes.
With the new solution, the systems should run more stable and scale flexibly with the performance requirements. In addition, the specifications included mirroring data, ensuring high availability and creating secure connections between the Deichmann and Oracle data centers.
The project was divided into two phases. The first phase included the migration of the productive environment of the data warehouse to the Exadata Cloud Service until the go-live in June 2021. The old system continued to operate the IT team in parallel as a backup. In the new environment, a number of processes were parallelized and container databases were set up. This required building up new know-how and adapting the source code in places. For the parallel operation of cloud and on-premises, the code base had to be kept uniform. In the second phase from summer 2021, the IT team moved the test and development environment to the cloud.
Despite the fact that data stocks are moved to Oracle’s data center via a virtual Private Network (VPN), the new system is two and a half hours faster than the old one in the nightly batch processes. This means that the system resources for day-to-day business are available earlier in the day without restrictions and without competition between parallel processes.
There have also been fewer helpdesk requests for reports that are too slow. The IT team uses the time gained to develop new applications and services. The new system also enables Deichmann to use business intelligence in day-to-day business. Previously, this had not been possible due to poor performance.
*Jens Dose is editor of the CIO magazine. In addition to the core topics around CIOs and their projects, he also deals with the role of the CISO and its area of responsibility.