Join processing heuristics in relational database systems

摘要

Due to its direct impact on the performance of relational database management systems (DBMS), query processing algorithms have been the subject of considerable study. Typical processing strategies involve the calculation of elaborate cost functions to find an “optimum” one. The topic of this paper is to develop optimization heuristics to complement (and sometimes replace) these elaborate cost functions with emphasis on the optimization of join queries. The heuristics are identified by means of detailed simulations of a particular DBMS model under various environmental conditions. The system model is based on System R.