Here are the steps that we need to take to make streaming loading possible for LucidDB:
Prerequisites:
- Make sure we run on a server version >= 0.7.5
- Create an empty FIFO directory
The following steps are carried out by the step at runtime:
- In the FIFO directory, create a FIFO file called {tableName}.csv (using mkfifo, LINUX ONLY FOLKS!)
- Create a target table using standard Kettle SQL generation
- Create a fifo server (with a certain name) on LucidDB