Request-driven architecture uses synchronous communication where an orchestrator service manages the flow between other services.
Event-driven architecture uses an event broker to decouple services, allowing them to communicate asynchronously through events.
Event-driven systems offer resilience and plug-and-play functionality but introduce challenges like unclear process flows and complex debugging.
The event notification pattern sends a minimal event with just an ID, requiring consumers to make a callback for more data, which favors consistency over availability.
The event-carried state transfer pattern includes all necessary data within the event itself, eliminating callbacks and promoting service autonomy and availability.
An event store captures the complete history of all events, enabling advanced analytics and answering questions that current-state databases cannot.
Event sourcing uses the event store as the single source of truth, allowing systems to be rebuilt or migrated by replaying events from the beginning.
An autonomous microservice is highly independent, prioritizes availability over consistency, and uses events for communication to achieve resilience.
The primary goal is to deliver business value efficiently, which means choosing the simplest architecture that works, even if it's request-driven.
26:39Loïc Carbonne
27:19Pratishtha Pandey
46:24David Leitner
25:05Eberhard Wolff
43:27Allard Buijze
22:19Abhimanyu Selvan
59:03Florian Rappl
26:08Maxim Fateev
.gif?w=240&auto=compress,format)
Jobs that call for the skills explored in this talk.
Peter Park System GmbH
München, Germany
CONTIAMO GMBH
Berlin, Germany
M&M Software GmbH
Sankt Georgen im Schwarzwald, Germany
doinstruct Software GmbH
Berlin, Germany
