Enabling Operator Reordering in Data Flow Programs Through Static Code Analysis

Abstract

In many massively parallel data management platforms, programs are represented as small imperative pieces of code connected in a data flow. This popular abstraction makes it hard to apply algebraic reordering techniques employed by relational DBMSs and other systems that use an algebraic programming abstraction. We present a code analysis technique based on reverse data and control flow analysis that discovers a set of properties from user code, which can be used to emulate algebraic optimizations in this setting.