Hurence / logisland

Scalable stream processing platform for advanced realtime analytics on top of Kafka and Spark. LogIsland also supports MQTT and Kafka Streams (Flink being in the roadmap). The platform does complex event processing and is suitable for time series analysis. A large set of valuable ready to use processors, data sources and sinks are available.

Home Page:https://logisland.github.io

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Support the instanciation of pipelines - have a proper pipeline loader

lhubert opened this issue · comments

Expected behavior and actual behavior.

We want to be able to run pipelines in multiple environments (Spark, Flink, Java bare metal in an application). At the moment the pipelines use class loaders that make instantiations difficult. We should introduce the concept of pipeline and provide a loader that give us more flexibility in running our pipelines in different situations.

Steps to reproduce the problem.

Specifications like the version of the project, operating system, or hardware.