{REAL-TIME DATA STREAMING & MONITORING INFRASTRUCTURE}
Bim
Oredata and BİM collaborated to implement a new Confluent Kafka cluster and enhance monitoring and alerting mechanisms. This project aimed to optimize data flow management between BİM’s branches, improve system resialience, and modernize monitoring tools to ensure greater efficiency and stability across all environments.
ABOUT BİM
BİM Birleşik Mağazalar A.Ş., the company with the highest market share in the organized retail sector in Türkiye, began its operations with 21 stores in 1995. BİM’s core principle is to offer consumers basic food items and consumer goods by combining high quality with the most affordable prices. As the pioneer of the hard-discount model in Türkiye, BİM’s product portfolio includes around 900 items, aiming to provide a large range of private label products to its customers.
{PRODUCTS WE USED}
The Challenge
BİM was managing inbound and outbound data from its branches through a single Kafka cluster and wanted to separate this traffic. Additionally, the monitoring tools in their environments were outdated, and the disaster recovery clusters lacked a robust monitoring structure.
The Solution
As a solution, we created a new Confluent Kafka cluster using Ansible. A playbook was designed and implemented according to the customer’s requirements. For the monitoring part with Grafana and Prometheus, we first planned the resources for the servers running these services based on growth, and then updated both applications.
The Result
As a result, BİM’s inbound and outbound traffic was successfully separated, and a second production cluster was installed. In addition to the new environment, metrics for the disaster recovery cluster also began to be collected. All three clusters could now be monitored end-to-end using the latest version of Grafana, alarm mechanisms were established across all environments, and a resilient structure was achieved.
