System R is an experimental data base management system that was designed t
o be unusually easy to use. System R supports a high-level relational user
language called SQL, which may be used by ad hoc users at terminals or by p
rogrammers as an imbedded data sublanguage in PL/I or COBOL. This paper des
cribes the overall architecture of the system, including the Relational Dat
a System (RDS) and the Research Storage System (RSS).
RDS is a data base language compiler. Host language programs with imbedded
SQL, statements are compiled by System R, which replaces the SQL. statement
s with calls to a machine-language access module. The compilation approach
removes much of the work of parsing, name binding, and optimization from th
e path of a running program, enabling highly efficient support for repetiti
ve transactions. In contrast, the RSS is a low-level DBMS, supporting simpl
e record-at-a-time operators, but with rather sophisticated transaction man
agement, recovery, and concurrency control.