The Channel Holder service then starts the asynchronous flow. The data enrichment helps the backend decide if the solution should use a legacy microservice system or a new one to process the EFT. The solution validates this initial request using the EFT Pilot Password service.īesides carrying out validations, the service also enriches the data. The service provides synchronous responses to requesters using a publish-subscribe pattern through an Azure Cache for Redis and waits for a backend response. WorkflowĪt the core, the backend services provide the necessary logic for an EFT to happen:Ī new EFT starts with an HTTP request received by the Channel Holder service. The CSE team created the other microservices as stubs to specifically isolate the actual Contoso microservices from other external mainframe services that the solution pushed through Azure Pipelines. Scalability and Performance for Transaction Simulation deliverables.Īzure Kubernetes Services (AKS) for the node autoscaler for Channel Holder. This cluster was:Īzure Red Hat OpenShift (ARO) in the Kubernetes/OpenShift Horizontal Pod Autoscaler (HPA) for: The actual Contoso microservices containers were manually pushed through Docker to the Kubernetes cluster.
Three main blocks make up the solution: back-end services, load testing, and monitoring with Event Autoscaler. Architectureĭownload a Visio file of this architecture. It's a major international financial services industry (FSI) organization that wanted to modernize one of its financial transaction systems. For the sake of anonymity, the article refers to the customer as Contoso Bank. This article summarizes the process and components the Microsoft Commercial Software Engineering (CSE) team used to build a solution for a banking customer.
0 kommentar(er)
