AsterixDB currently performs interval joins but it is not optimized; the database uses a nested loop when an interval join is required. In the process of learning more about Asterix I was shown an out-of-date, never merged branch that implemented an optimized version of interval join. The goal of my project is to update and migrate that interval join code into the active repository. The first part of my project will be to merge the interval join code into my own branch, update it, and update the interval optimization rules. The second part of the project is to update the query parser so that Asterix will run an interval join using the new query plan. The last stage of my plan is to update/create any tests that need to be modified, and perform full scale testing.


Caleb Herbel


  • prestonc@apache