（net time : 92 min)

Section 3 Database

DBMS = API + query interface + admin interface + programs

database = physical data store + schema

schema= organization + association + data spec + data control

data spec: type, length, location, indexing

data control: validation, dependency, access rights

Benefits: simultaneous access, simple query language, uniform and consistent access and content control, support distributed data

natural key: atomic weight

steps: table -> primary key -> 1 to n(foreign key) or n to n(table)

              -> classification hierarchy (union or not)->constraints-> evaluate schema quality (flexible, no redundancy)

              ->data type -> integrity and security control

1NF: same number of columns; 2NF: all functionally dependent on primary key; 3NF: all functionally independent on any other non-key attribute.

functional dependency: 1 on 1,

architecture: single (DBMS + server, simple), replicated (multiple, fault-tolerant), partitioned (single + multiple, minimize sync), cloud (cheap, expert)

risk factors: architecture(high), existing databases(high), domain model class diagram(low)

              high risk -> designed in the first few iteration

integrity control objectives: do the right thing(authorization) right(error-proof), security(confidential data)

transaction logging: discourage malicious changes, recovery mechanism

4 input controls: value limit, completeness, validation, combination

security control objectives: stable and functioning operating environment, protecting info and transaction in insecure environments

              certifying authorities: axiom

              (covered in operating system already)