case study
Urban Outfitters
Founded
1970
Headquarters
USA
Industry
Retail
Number of Locations
260+
Abstract
Urban Outfitters was launching a large enterprise-wide order management system upgrade which included the migration of their On-Premise Sterling OMS from v9 to v10. This upgrade included migration from Oracle to Postgres DB. Urban was looking for Sterling OMS Performance expertise to help with a complete performance assessment to ensure the newly upgraded system would perform comparably with the existing OMS solution.
About UrbanOutfitters
Urban Outfitters is a global lifestyle retailer known for its bohemian-inspired apparel, accessories, and home goods.
The company operates a portfolio of brands including Urban Outfitters, Anthropologie, Free People, and BHLDN, catering to diverse customer segments. With a significant online presence and a growing international footprint, Urban Outfitters is a leader when it comes to omnichannel commerce excellence in the specialty retail industry.
IBM Sterling OMS Performance Testing Execution
-
Urban Outfitters is one of IBM’s longest tenured customers on the IBM Order Management platform and they recently undertook a massive upgrade effort to move everything from v9 on premise to v10 of IBM Sterling OMS using Containers on Google Cloud Platform (GCP), which included a migration from Oracle DB to PostrgeSQL. The objective of the Perfaware team was to ensure that the system solution meets the performance requirements, as specified, and could effectively support the required number of users and planned volumes of transaction data post migration to the new architecture and technical stack. Additionally, the performance tests were designed to validate that the solution will scale on the technical and application architecture to meet anticipated future growth. Urban Outfitters were seeking full confidence that, once deployed, the solution would meet expectations as it relates to scalability, user responsiveness, and ease of support.
When it comes to scalability, it was imperative that the solution could scale to meet the peak hour and peak day transaction volumes and sustain them over a period of time. While under full load it was also important to ensure that the synchronous interactions with the system via the various user interfaces including users at their Stores or Call centers would meet response time goals. Finally the system had to show that it could operate over long periods with no SEV1 or SEV2 issues arising. Perfaware was able to deliver complete performance analysis and tuning of the IBM OMS on GCP in just under four months which allowed Urban Outfitters to go into the 2024 Holiday season with confidence that the new system would scale during their heaviest seasonal period.
Business Challenge
- Perfaware needed to be flexible given Urban’s migration from on-prem to cloud was delayed and now without some troubles. Some components could not be tested as they were not available at the moment.
- Perfaware had to carefully navigate the constant scope creep that Urban requested. Despite no clear scope definition being provided initially, Perfaware was able to come up with a proposed scope to address their most pressing needs and this only required a couple of changes along the way to fully deliver the expected results.
- The NFRs and test scenarios were corrected as the test was going on. Luckily, Perfaware had the necessary skills, time and flexibility to prevent these issues from impacting the cost to the customer..
- There were periodic access issues, changes in the topology and infrastructure.
- Perfaware had to adapt to the lack of adequate testing of core components in the infrastructure such as JMS, AlloyDB, and Kafka and helped to stabilize these components without impacting the cost to the customer.
Getting ready for holiday peak on a new tech stack
Results
- The main volume-related and many high risk scenarios were load tested with enough volume and variety to ensure they’ll perform in acceptable manner during the peak load.
- With close collaboration between Google, IBM and the client, the Perfaware team was able to identify and tune noted bottlenecks and hot spots. A few issues needed to be or bypassed, or documented well enough to optimize them outside of the project scope.
- The Urban team received the optimized Performance environment with the necessary software, settings and scripts as well as multiple sessions of knowledge transfer to enable them to be self-sufficient for executing future load tests.
Fill in the form below
To know more about the implementation