Summary: How and why a leading manufacturing organization migrated from monolithic apps—to microservices—and then to service-mesh.
Problem: As we started cloud native transformation to host microservices on Pivotal Cloud Foundry (now part of VMware Tanzu), challenges of managing multiple services on cloud became clear (Death Star). Also, no single product provided all the capabilities (context-sensitive circuit breaker).
Solution: We didn’t want to create a DIY service mesh solution, thrown away once the platform supports it. With Pivotal’s guidance, we developed a service orchestration solution by leveraging frameworks like Zuul, Eureka, and Spring Cloud Data Flow. It serves as a dedicated layer for svc-to-svc communication and provides multitenancy, enabling global deployments of apps.
Conclusion: Teams were creating one-off solutions for capabilities like multitenancy, discovery, and retry. “Transition architecture” resulted in significant cost savings, reduced time-to-value, and more importantly, secured microservices.
Senior IT Architect, GM
Track: Architecture