Twister2 is a big data tool kit supporting different parallel computing paradigms for big data applications. Twister2 models a parallel computation as a graph. Depending on the requirements the graph can be deployed, executed with different choices making Twister2 capable of running different applications including classic parallel computing, data parallel, streaming or function as a service applications.
The detailed desing of Twister2 can be found at
We use Jira for tracking the progress of twister2.
The google groups mailing list can be used to ask questions and discussions "twister2@googlegroups.com"