Nikita Salnikov-Tarnovski

The majority of learning materials about streaming data processing is focused on code examples of how to use the specific tool or framework. But to have working system in production you need much more than just business logic. Where does your data come from? How long do you keep itaround? How do you upgrade your application and handle data schema changes? In this talk I will talk about operation concerns of streaming applications heavily leaning on our own experience in Plumbr. We process billions of messages per month for several years now and this gave us a nice perspective of what is important and what is not so much.