DITA: Distributed In-Memory Trajectory Analytics

DITA is a distributed in-memory trajectory analytics system based on Apache Spark 2.2.0.

Development

Since we use IntelliJ for development, you can consult the official guide for setuping up the IDE. Besides, you should do the following things:

Go to View > Tool Windows > Maven Projects and add hadoop-2.6, hive-provided, hive-thriftserver, yarn in Profiles (there are some default profiles as well, don't change them). Then Reimport All Maven Projects (the first button on upper-right corner), Generate Sources and Update Folders For All Projects (the second button on upper-right corner).
Rebuild the whole project, which would fail but is essential for following steps.
Marking Generated Sources:
- Go to File > Project Structure > Project Settings > Modules. Find spark-streaming-flume-sink, and mark target/scala-2.11/src_managed/main/compiled_avro as source. (Click on the Sources on the top to mark)
- Go to File > Project Structure > Project Settings > Modules. Find spark-hive-thriftserver, and mark src/gen/java as source. (Click on the Sources on the top to mark)
Rebuild the whole project again, which should work well now. If there still exist some compilation errors for not finding some classes, you may return to last step and marking corresponding sources if not included.

The master branch is the version integrated with Spark SQL, and the standalone branch is a stand-alone version just with DITA code.