A leading US grocery retailer wanted to streamline their queue management processes, reduce technical debt, and achieve substantial cost savings.
The problem.
The retailer faced numerous challenges, including a small pool of administrators coupled with a continually shrinking budget, and a lack of expertise in maintaining their stable mainframe. Queue Manager restarts were cumbersome and problematic, often causing missed fulfillment cutoff times and resulting in stockouts. Additionally, MQ disruptions every Wednesday severely impacted the promotional engine, finance, and recording functions, further exacerbating operational difficulties.
Background.
- Leading US Grocery retailer
- Combined MQ, Mainframe, and Apache Kafka® installation distributed across 2,400 retail stores
- 24 administrators govern performance, but Engineers, Application Teams & DevOps professionals need access to MQ Queues
- Key objective of reducing technical debt & YOY IT spend
- Coupons are pushed to mobile apps from stores every Wednesday
- Moving the mainframe to the Kyndryl cloud as part of a larger cloud transformation initiative
Challenges.
- A small pool of administrators coupled with an ever-shrinking budget. The Mainframe is stable & not going anywhere, but the lack of expertise is a continual challenge
- Restarting Queue Managers was cumbersome, problematic, and regularly caused fulfillment cutoff times to be missed, resulting in stockouts
- MQ disruption on Wednesday causes promotional engine, finance & recording to be severely impacted
Capabilities.
- Single tool & login for all MQ & Mainframe instances for all management & configuration needs
- Al/ML monitoring to predict & alert outages
- “Service Layer” created between MQ & Grafana for usage by a separate Observability team
Final outcomes.
- All DevOps Engineers have access to MQ & Apache Kafka® monitors & can restart or shutdown all based on their needs
- 33% – 47% Increase in Administrative Efficiency & YoY Support Cost Reductions
- Zero MQ & Mainframe Disruptions