The open-closed principle, fault tolerance, technology diversity, and independent deployment are essential goals for a robust architecture.
Microservices decompose the business and persistence layers into small, independent services, each with its own data store.
Sharing interfaces or using a common parent POM creates tight coupling that breaks independent development and builds.
Common functionality should be outsourced to versioned shared libraries, and service APIs must be versioned to prevent breaking changes.
Using multiple communication channels for the same data object complicates maintenance, so a single entry point simplifies logic and conversion.
Reduce the impact of service failures caused by temporal coupling by implementing backup strategies like default values or disabling features.
Implementing a local data cache, such as Google Guava, reduces synchronous calls to other services and improves system resilience.
Use platforms like Kubernetes with health checks to automatically scale and replace faulty instances of stateless services for high availability.
Asynchronous communication via a message broker increases availability and fault tolerance but requires careful handling of redelivery and monitoring.
Building a microservice architecture is an ongoing, iterative process that requires continuous improvement and an open-minded team culture.
1:02:46Aleksandr Kalikov
1:07:36Josh Long
49:30Moritz Kammerer
54:18Michael Eder & Philipp Frauenthaler
57:38Bernd Ruecker
50:06Jens Happe
52:24Lukas Georgieff & Alberto Gisbert
54:30Philipp Frauenthaler & Gregor Lucny
.gif?w=240&auto=compress,format)
Jobs that call for the skills explored in this talk.
BMS Corporate Solutions GmbH
Düsseldorf, Germany
IT-Objects GmbH
Essen, Germany
Saby New Compy
Bielefeld, Germany
B & B Unternehmensberatung GmbH & Co. KG