Tweet
Tweet

Slide 1

Slide 1 text

Efficient Mutation Analysis of Relational Database Structure Using Mutant Schemata and Parallelisation Chris J. Wright Gregory M. Kapfhammer Phil McMinn Mutation 2013

Slide 2

Slide 2 text

Chris J. Wright - [email protected] Relational Database Management Systems (RDBMS)

Slide 3

Slide 3 text

Chris J. Wright - [email protected] Relational Database Management Systems (RDBMS)

Slide 4

Slide 4 text

Chris J. Wright - [email protected]

Slide 5

Slide 5 text

Chris J. Wright - [email protected] Many different RDBMSs...

Slide 6

Slide 6 text

Chris J. Wright - [email protected] Many different RDBMSs... ...same specification of structure

Slide 7

Slide 7 text

Chris J. Wright - [email protected] 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 9 REFERENCES T (A, B) 10 ); Database Schema

Slide 8

Slide 8 text

Chris J. Wright - [email protected] 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 9 REFERENCES T (A, B) 10 ); Database Schema Tables

Slide 9

Slide 9 text

Chris J. Wright - [email protected] 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 9 REFERENCES T (A, B) 10 ); Database Schema Columns

Slide 10

Slide 10 text

Chris J. Wright - [email protected] 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 9 REFERENCES T (A, B) 10 ); Database Schema Constraints

Slide 11

Slide 11 text

Chris J. Wright - [email protected] 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 9 REFERENCES T (A, B) 10 ); Database Schema How do we know this is correct?

Slide 12

Slide 12 text

Chris J. Wright - [email protected] Why Test Database Structure?

Slide 13

Slide 13 text

Chris J. Wright - [email protected] Database Schema Why Test Database Structure?

Slide 14

Slide 14 text

Chris J. Wright - [email protected] DBMS Database Schema Why Test Database Structure?

Slide 15

Slide 15 text

Chris J. Wright - [email protected] DBMS Database Schema Why Test Database Structure?

Slide 16

Slide 16 text

Chris J. Wright - [email protected] DBMS Application Database Schema Why Test Database Structure?

Slide 17

Slide 17 text

Chris J. Wright - [email protected] DBMS Application Web Server Database Schema Why Test Database Structure?

Slide 18

Slide 18 text

Chris J. Wright - [email protected] DBMS Application Web Server Third Party Database Schema Why Test Database Structure?

Slide 19

Slide 19 text

Chris J. Wright - [email protected] DBMS Application Web Server Third Party Database Schema ✗ Why Test Database Structure?

Slide 20

Slide 20 text

Chris J. Wright - [email protected] DBMS Application Web Server Third Party Database Schema ✗ ✗ Why Test Database Structure?

Slide 21

Slide 21 text

Chris J. Wright - [email protected] DBMS Application Web Server Third Party Database Schema ✗ ✗ ✗ Why Test Database Structure?

Slide 22

Slide 22 text

Chris J. Wright - [email protected] DBMS Application Web Server Third Party Database Schema ✗ ✗ ✗ ✗ Why Test Database Structure?

Slide 23

Slide 23 text

Chris J. Wright - [email protected] Mutation Analysis

Slide 24

Slide 24 text

Chris J. Wright - [email protected] Mutation Analysis Test Suite Application

Slide 25

Slide 25 text

Chris J. Wright - [email protected] Mutation Analysis Database Test Suite

Slide 26

Slide 26 text

Chris J. Wright - [email protected] Mutation Analysis Database Insert Statements

Slide 27

Slide 27 text

Chris J. Wright - [email protected] Mutation Analysis Database Insert Statements

Slide 28

Slide 28 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 9 REFERENCES T (A, B) 10 ); Database Insert Statements

Slide 29

Slide 29 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 9 REFERENCES T (A, B) 10 ); Database Insert Statements

Slide 30

Slide 30 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 9 REFERENCES T (A, B) 10 ); INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); Database Insert Statements

Slide 31

Slide 31 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 9 REFERENCES T (A, B) 10 ); INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✓ Database Insert Statements

Slide 32

Slide 32 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 9 REFERENCES T (A, B) 10 ); INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✓ Database Insert Statements

Slide 33

Slide 33 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 9 REFERENCES T (A, B) 10 ); INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✓ INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); Database Insert Statements

Slide 34

Slide 34 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 9 REFERENCES T (A, B) 10 ); INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✓ INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✗ Database Insert Statements

Slide 35

Slide 35 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 9 REFERENCES T (A, B) 10 ); INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✓ INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✗ Database Insert Statements

Slide 36

Slide 36 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 9 REFERENCES T (A, B) 10 ); INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✓ INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✗ INSERT INTO S(X, Y, Z) VALUES('a', 'a', 'b'); Database Insert Statements

Slide 37

Slide 37 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 9 REFERENCES T (A, B) 10 ); INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✓ INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✗ INSERT INTO S(X, Y, Z) VALUES('a', 'a', 'b'); ✓ Database Insert Statements

Slide 38

Slide 38 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 9 REFERENCES T (A, B) 10 ); INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✓ INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✗ INSERT INTO S(X, Y, Z) VALUES('a', 'a', 'b'); ✓ Database Insert Statements

Slide 39

Slide 39 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 9 REFERENCES T (A, B) 10 ); INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✓ INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✗ INSERT INTO S(X, Y, Z) VALUES('a', 'a', 'b'); ✓ Database Insert Statements INSERT INTO S(X, Y, Z) VALUES('a', 'b', 'a');

Slide 40

Slide 40 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 9 REFERENCES T (A, B) 10 ); INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✓ INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✗ INSERT INTO S(X, Y, Z) VALUES('a', 'a', 'b'); ✓ Database Insert Statements INSERT INTO S(X, Y, Z) VALUES('a', 'b', 'a'); ✗

Slide 41

Slide 41 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 9 REFERENCES T (A, B) 10 ); INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✓ INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✗ INSERT INTO S(X, Y, Z) VALUES('a', 'a', 'b'); ✓ Database Insert Statements INSERT INTO S(X, Y, Z) VALUES('a', 'b', 'a'); ✗

Slide 42

Slide 42 text

Chris J. Wright - [email protected] Mutating the Structure 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, ) 9 REFERENCES T (A, B) 10 ); Y Z

Slide 43

Slide 43 text

Chris J. Wright - [email protected] Mutating the Structure 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, ) 9 REFERENCES T (A, B) 10 ); Y Z

Slide 44

Slide 44 text

Chris J. Wright - [email protected] Mutating the Structure 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, ) 9 REFERENCES T (A, B) 10 ); Y Z

Slide 45

Slide 45 text

Chris J. Wright - [email protected] Mutating the Structure 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, ) 9 REFERENCES T (A, B) 10 ); Y Z

Slide 46

Slide 46 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 );

Slide 47

Slide 47 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 );

Slide 48

Slide 48 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); Database (mutated)

Slide 49

Slide 49 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); Database (mutated) Insert Statements

Slide 50

Slide 50 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); Database (mutated) Insert Statements

Slide 51

Slide 51 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✓ Database (mutated) Insert Statements

Slide 52

Slide 52 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✓ Database (mutated) Insert Statements

Slide 53

Slide 53 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✓ INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); Database (mutated) Insert Statements

Slide 54

Slide 54 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✓ INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✗ Database (mutated) Insert Statements

Slide 55

Slide 55 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✓ INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✗ Database (mutated) Insert Statements

Slide 56

Slide 56 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✓ INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✗ INSERT INTO S(X, Y, Z) VALUES('a', 'a', 'b'); Database (mutated) Insert Statements

Slide 57

Slide 57 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✓ INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✗ INSERT INTO S(X, Y, Z) VALUES('a', 'a', 'b'); ✓ Database (mutated) Insert Statements

Slide 58

Slide 58 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✓ INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✗ INSERT INTO S(X, Y, Z) VALUES('a', 'a', 'b'); ✓ Database (mutated) Insert Statements

Slide 59

Slide 59 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✓ INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✗ INSERT INTO S(X, Y, Z) VALUES('a', 'a', 'b'); ✓ Database (mutated) Insert Statements INSERT INTO S(X, Y, Z) VALUES('a', 'b', 'a');

Slide 60

Slide 60 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✓ INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✗ INSERT INTO S(X, Y, Z) VALUES('a', 'a', 'b'); ✓ Database (mutated) Insert Statements INSERT INTO S(X, Y, Z) VALUES('a', 'b', 'a'); ✓

Slide 61

Slide 61 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✓ INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✗ INSERT INTO S(X, Y, Z) VALUES('a', 'a', 'b'); ✓ Database (mutated) Insert Statements INSERT INTO S(X, Y, Z) VALUES('a', 'b', 'a'); ✓

Slide 62

Slide 62 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✓ INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✗ INSERT INTO S(X, Y, Z) VALUES('a', 'a', 'b'); ✓ Database (mutated) Insert Statements INSERT INTO S(X, Y, Z) VALUES('a', 'b', 'a'); ✓ Results are different

Slide 63

Slide 63 text

Chris J. Wright - [email protected] Mutation Analysis 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✓ INSERT INTO T(A, B, C) VALUES('a', 'a', 'a'); ✗ INSERT INTO S(X, Y, Z) VALUES('a', 'a', 'b'); ✓ Database (mutated) Insert Statements INSERT INTO S(X, Y, Z) VALUES('a', 'b', 'a'); ✓ Mutant is killed

Slide 64

Slide 64 text

Chris J. Wright - [email protected] Mutation Operators

Slide 65

Slide 65 text

Chris J. Wright - [email protected] Mutation Operators Primary Key

Slide 66

Slide 66 text

Chris J. Wright - [email protected] Mutation Operators Primary Key Not Null

Slide 67

Slide 67 text

Chris J. Wright - [email protected] Mutation Operators Check Primary Key Not Null

Slide 68

Slide 68 text

Chris J. Wright - [email protected] Mutation Operators Check Primary Key Not Null Unique

Slide 69

Slide 69 text

Chris J. Wright - [email protected] 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); The Problem? 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 );

Slide 70

Slide 70 text

Chris J. Wright - [email protected] 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); The Problem? 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 );

Slide 71

Slide 71 text

Chris J. Wright - [email protected] 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); The Problem? Many mutants to analyse

Slide 72

Slide 72 text

Chris J. Wright - [email protected] 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); The Problem? A time consuming process

Slide 73

Slide 73 text

Chris J. Wright - [email protected] The Solution?

Slide 74

Slide 74 text

Chris J. Wright - [email protected] The Solution? Mutant Schemata

Slide 75

Slide 75 text

Chris J. Wright - [email protected] The Solution? Mutant Schemata Combine mutants into a ‘meta-mutant’

Slide 76

Slide 76 text

Chris J. Wright - [email protected] The Solution? Parallelisation Mutant Schemata Combine mutants into a ‘meta-mutant’

Slide 77

Slide 77 text

Chris J. Wright - [email protected] The Solution? Parallelisation Mutant Schemata Combine mutants into a ‘meta-mutant’ Analyse multiple mutants simultaneously

Slide 78

Slide 78 text

Chris J. Wright - [email protected] Mutation Analysis Approaches Mutant Representation Parallelisation Strategy Normal “Original”

Slide 79

Slide 79 text

Chris J. Wright - [email protected] Mutation Analysis Approaches Mutant Representation Parallelisation Strategy Schemata Normal “Original” Full “Full Schemata”

Slide 80

Slide 80 text

Chris J. Wright - [email protected] Original Approach

Slide 81

Slide 81 text

Chris J. Wright - [email protected] Original Approach Create structure in database

Slide 82

Slide 82 text

Chris J. Wright - [email protected] Original Approach Create structure in database Execute insert statements

Slide 83

Slide 83 text

Chris J. Wright - [email protected] Original Approach Create structure in database Execute insert statements Drop structure from database

Slide 84

Slide 84 text

Chris J. Wright - [email protected] Original Approach Create structure in database Execute insert statements Drop structure from database

Slide 85

Slide 85 text

Chris J. Wright - [email protected] Mutant Schemata Approach Create structure in database Execute insert statements Drop structure from database

Slide 86

Slide 86 text

Chris J. Wright - [email protected] Mutant Schemata Approach Create structure in database Execute insert statements Drop structure from database

Slide 87

Slide 87 text

Chris J. Wright - [email protected] Mutant Schemata Approach Create structure in database Execute insert statements Drop structure from database Reduce time creating/ dropping database

Slide 88

Slide 88 text

Chris J. Wright - [email protected] Full Schemata Approach

Slide 89

Slide 89 text

Chris J. Wright - [email protected] Full Schemata Approach 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 );

Slide 90

Slide 90 text

Chris J. Wright - [email protected] Full Schemata Approach 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES T (A, B) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, C) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 9 REFERENCES T (A, B) 10 );

Slide 91

Slide 91 text

Chris J. Wright - [email protected] Full Schemata Approach 1 CREATE TABLE mutant_1_T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE mutant_1_S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES mutant_1_T (A, B) 10 ); 1 CREATE TABLE mutant_2_T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, C) 4 ); 5 6 CREATE TABLE mutant_2_S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 9 REFERENCES mutant_2_T (A, B) 10 );

Slide 92

Slide 92 text

Chris J. Wright - [email protected] Full Schemata Approach 1 CREATE TABLE mutant_1_T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE mutant_1_S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES mutant_1_T (A, B) 10 ); 11 12 CREATE TABLE mutant_2_T ( 13 A CHAR, B CHAR, C CHAR, 14 CONSTRAINT UniqueOnColsAandB UNIQUE (A, C) 15 ); 16 17 CREATE TABLE mutant_2_S ( 18 X CHAR, Y CHAR, Z CHAR, 19 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 20 REFERENCES mutant_2_T (A, B) 21 );

Slide 93

Slide 93 text

Chris J. Wright - [email protected] Full Schemata Approach 1 CREATE TABLE mutant_1_T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE mutant_1_S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES mutant_1_T (A, B) 10 ); 11 12 CREATE TABLE mutant_2_T ( 13 A CHAR, B CHAR, C CHAR, 14 CONSTRAINT UniqueOnColsAandB UNIQUE (A, C) 15 ); 16 17 CREATE TABLE mutant_2_S ( 18 X CHAR, Y CHAR, Z CHAR, 19 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 20 REFERENCES mutant_2_T (A, B) 21 );

Slide 94

Slide 94 text

Chris J. Wright - [email protected] Full Schemata Approach 1 CREATE TABLE mutant_1_T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE mutant_1_S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES mutant_1_T (A, B) 10 ); 11 12 CREATE TABLE mutant_2_T ( 13 A CHAR, B CHAR, C CHAR, 14 CONSTRAINT UniqueOnColsAandB UNIQUE (A, C) 15 ); 16 17 CREATE TABLE mutant_2_S ( 18 X CHAR, Y CHAR, Z CHAR, 19 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 20 REFERENCES mutant_2_T (A, B) 21 ); Database (mutated) Insert Statements

Slide 95

Slide 95 text

Chris J. Wright - [email protected] Full Schemata Approach 1 CREATE TABLE mutant_1_T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE mutant_1_S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES mutant_1_T (A, B) 10 ); 11 12 CREATE TABLE mutant_2_T ( 13 A CHAR, B CHAR, C CHAR, 14 CONSTRAINT UniqueOnColsAandB UNIQUE (A, C) 15 ); 16 17 CREATE TABLE mutant_2_S ( 18 X CHAR, Y CHAR, Z CHAR, 19 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 20 REFERENCES mutant_2_T (A, B) 21 ); Database (mutated) INSERT INTO S(X, Y, Z) VALUES('a', 'a', 'a'); Insert Statements

Slide 96

Slide 96 text

Chris J. Wright - [email protected] Full Schemata Approach 1 CREATE TABLE mutant_1_T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE mutant_1_S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES mutant_1_T (A, B) 10 ); 11 12 CREATE TABLE mutant_2_T ( 13 A CHAR, B CHAR, C CHAR, 14 CONSTRAINT UniqueOnColsAandB UNIQUE (A, C) 15 ); 16 17 CREATE TABLE mutant_2_S ( 18 X CHAR, Y CHAR, Z CHAR, 19 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 20 REFERENCES mutant_2_T (A, B) 21 ); Database (mutated) INSERT INTO mutant_1_S(X, Y, Z) VALUES('a', 'a', 'a'); Insert Statements

Slide 97

Slide 97 text

Chris J. Wright - [email protected] Full Schemata Approach 1 CREATE TABLE mutant_1_T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE mutant_1_S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES mutant_1_T (A, B) 10 ); 11 12 CREATE TABLE mutant_2_T ( 13 A CHAR, B CHAR, C CHAR, 14 CONSTRAINT UniqueOnColsAandB UNIQUE (A, C) 15 ); 16 17 CREATE TABLE mutant_2_S ( 18 X CHAR, Y CHAR, Z CHAR, 19 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 20 REFERENCES mutant_2_T (A, B) 21 ); Database (mutated) INSERT INTO mutant_1_S(X, Y, Z) VALUES('a', 'a', 'a'); ✓ Insert Statements

Slide 98

Slide 98 text

Chris J. Wright - [email protected] Full Schemata Approach 1 CREATE TABLE mutant_1_T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE mutant_1_S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES mutant_1_T (A, B) 10 ); 11 12 CREATE TABLE mutant_2_T ( 13 A CHAR, B CHAR, C CHAR, 14 CONSTRAINT UniqueOnColsAandB UNIQUE (A, C) 15 ); 16 17 CREATE TABLE mutant_2_S ( 18 X CHAR, Y CHAR, Z CHAR, 19 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 20 REFERENCES mutant_2_T (A, B) 21 ); Database (mutated) INSERT INTO mutant_1_S(X, Y, Z) VALUES('a', 'a', 'a'); ✓ Insert Statements

Slide 99

Slide 99 text

Chris J. Wright - [email protected] Full Schemata Approach

Slide 100

Slide 100 text

Chris J. Wright - [email protected] Create tables once... Full Schemata Approach

Slide 101

Slide 101 text

Chris J. Wright - [email protected] Create tables once... ...rewrite queries to match Full Schemata Approach

Slide 102

Slide 102 text

Chris J. Wright - [email protected] Mutation Analysis Approaches Mutant Representation Parallelisation Strategy Schemata Normal “Original” Full “Full Schemata”

Slide 103

Slide 103 text

Chris J. Wright - [email protected] Mutation Analysis Approaches Mutant Representation Parallelisation Strategy Schemata Normal “Original” Full “Full Schemata” Minimal “Minimal Schemata”

Slide 104

Slide 104 text

Chris J. Wright - [email protected] Full Schemata Approach 1 CREATE TABLE mutant_1_T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE mutant_1_S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES mutant_1_T (A, B) 10 ); 11 12 CREATE TABLE mutant_2_T ( 13 A CHAR, B CHAR, C CHAR, 14 CONSTRAINT UniqueOnColsAandB UNIQUE (A, C) 15 ); 16 17 CREATE TABLE mutant_2_S ( 18 X CHAR, Y CHAR, Z CHAR, 19 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 20 REFERENCES mutant_2_T (A, B) 21 );

Slide 105

Slide 105 text

Chris J. Wright - [email protected] Full Schemata Approach 1 CREATE TABLE mutant_1_T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE mutant_1_S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 9 REFERENCES mutant_1_T (A, B) 10 ); 11 12 CREATE TABLE mutant_2_T ( 13 A CHAR, B CHAR, C CHAR, 14 CONSTRAINT UniqueOnColsAandB UNIQUE (A, C) 15 ); 16 17 CREATE TABLE mutant_2_S ( 18 X CHAR, Y CHAR, Z CHAR, 19 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 20 REFERENCES mutant_2_T (A, B) 21 ); Mutated Tables

Slide 106

Slide 106 text

Chris J. Wright - [email protected] Minimal Schemata Approach 1 CREATE TABLE mutant_1_S ( 2 X CHAR, Y CHAR, Z CHAR, 3 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 4 REFERENCES mutant_1_T (A, B) 5 ); 6 7 CREATE TABLE mutant_2_T ( 8 A CHAR, B CHAR, C CHAR, 9 CONSTRAINT UniqueOnColsAandB UNIQUE (A, C) 10 ); Mutated Tables

Slide 107

Slide 107 text

Chris J. Wright - [email protected] Minimal Schemata Approach 1 CREATE TABLE mutant_1_S ( 2 X CHAR, Y CHAR, Z CHAR, 3 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 4 REFERENCES mutant_1_T (A, B) 5 ); 6 7 CREATE TABLE mutant_2_T ( 8 A CHAR, B CHAR, C CHAR, 9 CONSTRAINT UniqueOnColsAandB UNIQUE (A, C) 10 );

Slide 108

Slide 108 text

Chris J. Wright - [email protected] Minimal Schemata Approach 1 CREATE TABLE mutant_1_S ( 2 X CHAR, Y CHAR, Z CHAR, 3 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 4 REFERENCES mutant_1_T (A, B) 5 ); 6 7 CREATE TABLE mutant_2_T ( 8 A CHAR, B CHAR, C CHAR, 9 CONSTRAINT UniqueOnColsAandB UNIQUE (A, C) 10 );

Slide 109

Slide 109 text

Chris J. Wright - [email protected] Minimal Schemata Approach 1 CREATE TABLE mutant_1_S ( 2 X CHAR, Y CHAR, Z CHAR, 3 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 4 REFERENCES mutant_1_T (A, B) 5 ); 6 7 CREATE TABLE mutant_2_T ( 8 A CHAR, B CHAR, C CHAR, 9 CONSTRAINT UniqueOnColsAandB UNIQUE (A, C) 10 );

Slide 110

Slide 110 text

Chris J. Wright - [email protected] Minimal Schemata Approach 1 CREATE TABLE mutant_1_S ( 2 X CHAR, Y CHAR, Z CHAR, 3 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 4 REFERENCES mutant_1_T (A, B) 5 ); 6 7 CREATE TABLE mutant_2_T ( 8 A CHAR, B CHAR, C CHAR, 9 CONSTRAINT UniqueOnColsAandB UNIQUE (A, C) 10 ); 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 9 REFERENCES T (A, B) 10 );

Slide 111

Slide 111 text

Chris J. Wright - [email protected] Minimal Schemata Approach 1 CREATE TABLE T ( 2 A CHAR, B CHAR, C CHAR, 3 CONSTRAINT UniqueOnColsAandB UNIQUE (A, B) 4 ); 5 6 CREATE TABLE S ( 7 X CHAR, Y CHAR, Z CHAR, 8 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Y) 9 REFERENCES T (A, B) 10 ); 11 12 CREATE TABLE mutant_1_S ( 13 X CHAR, Y CHAR, Z CHAR, 14 CONSTRAINT RefToColsAandB FOREIGN KEY (X, Z) 15 REFERENCES T (A, B) 16 ); 17 18 CREATE TABLE mutant_2_T ( 19 A CHAR, B CHAR, C CHAR, 20 CONSTRAINT UniqueOnColsAandB UNIQUE (A, C) 21 );

Slide 112

Slide 112 text

Chris J. Wright - [email protected] Minimal Schemata Approach

Slide 113

Slide 113 text

Chris J. Wright - [email protected] Create tables once... Minimal Schemata Approach

Slide 114

Slide 114 text

Chris J. Wright - [email protected] Create tables once... ...only mutated tables... Minimal Schemata Approach

Slide 115

Slide 115 text

Chris J. Wright - [email protected] Create tables once... ...only mutated tables... Minimal Schemata Approach ...reduce queries executed

Slide 116

Slide 116 text

Chris J. Wright - [email protected] Create tables once... ...only mutated tables... Minimal Schemata Approach ...reduce queries executed Plus one copy for foreign keys

Slide 117

Slide 117 text

Chris J. Wright - [email protected] Mutation Analysis Approaches Mutant Representation Parallelisation Strategy Schemata Normal “Original” Full “Full Schemata” Minimal “Minimal Schemata”

Slide 118

Slide 118 text

Chris J. Wright - [email protected] Mutation Analysis Approaches Mutant Representation Parallelisation Strategy Schemata Normal “Original” Full “Full Schemata” Minimal “Minimal Schemata” Up front “Up-Front Schemata”

Slide 119

Slide 119 text

Chris J. Wright - [email protected] Mutation Analysis Approaches Mutant Representation Parallelisation Strategy Schemata Normal “Original” Full “Full Schemata” Minimal “Minimal Schemata” Up front “Up-Front Schemata” Just in time “Just-in-Time Schemata”

Slide 120

Slide 120 text

Chris J. Wright - [email protected] Parallelisation ‘Just-in-Time Schemata’

Slide 121

Slide 121 text

Chris J. Wright - [email protected] Parallelisation ‘Just-in-Time Schemata’ Make the ‘Original’ approach parallel

Slide 122

Slide 122 text

Chris J. Wright - [email protected] Parallelisation ‘Up-Front Schemata’ ‘Just-in-Time Schemata’ Make the ‘Original’ approach parallel

Slide 123

Slide 123 text

Chris J. Wright - [email protected] Parallelisation ‘Up-Front Schemata’ ‘Just-in-Time Schemata’ Make the ‘Original’ approach parallel Make the ‘Full Schemata’ approach parallel

Slide 124

Slide 124 text

Chris J. Wright - [email protected] Original Approach Create structure in database Execute insert statements Drop structure from database

Slide 125

Slide 125 text

Chris J. Wright - [email protected] ‘Just-in-Time’ Approach Create structure in database Execute insert statements Drop structure from database Parallel

Slide 126

Slide 126 text

Chris J. Wright - [email protected] ‘Just-in-Time’ Approach Create structure in database Execute insert statements Drop structure from database Parallel

Slide 127

Slide 127 text

Chris J. Wright - [email protected] Mutant Schemata Approach Create structure in database Execute insert statements Drop structure from database

Slide 128

Slide 128 text

Chris J. Wright - [email protected] ‘Up-Front’ Approach Create structure in database Execute insert statements Drop structure from database Parallel

Slide 129

Slide 129 text

Chris J. Wright - [email protected] Empirical Study

Slide 130

Slide 130 text

Chris J. Wright - [email protected] Empirical Study Evaluation Metric Mutation Time Approaches 5 Case Studies 6 DBMSs 2 Repetitions 30

Slide 131

Slide 131 text

Chris J. Wright - [email protected] Empirical Study Evaluation Metric Mutation Time Approaches 5 Case Studies 6 DBMSs 2 Repetitions 30

Slide 132

Slide 132 text

Chris J. Wright - [email protected] Empirical Study Evaluation Metric Mutation Time Approaches 5 Case Studies 6 DBMSs 2 Repetitions 30

Slide 133

Slide 133 text

Chris J. Wright - [email protected] Empirical Study Evaluation Metric Mutation Time Approaches 5 Case Studies 6 DBMSs 2 Repetitions 30

Slide 134

Slide 134 text

Chris J. Wright - [email protected] Empirical Study Evaluation Metric Mutation Time Approaches 5 Case Studies 6 DBMSs 2 Repetitions 30

Slide 135

Slide 135 text

Chris J. Wright - [email protected] Empirical Study: Approaches Mutant Representation Parallelisation Strategy Schemata Normal “Original” Full “Full Schemata” Minimal “Minimal Schemata” Up front “Up-Front Schemata” Just in time “Just-in-Time Schemata”

Slide 136

Slide 136 text

Chris J. Wright - [email protected] Empirical Study: Case Studies Case Study Tables Columns Primary Keys Foreign Keys Unique Constraints Cloc 2 10 0 0 0 JWhoisServer 6 49 6 0 0 NistDML182 2 32 1 1 0 NistDML183 2 6 0 1 1 RiskIt 13 56 11 10 0 UnixUsage 8 32 7 7 0

Slide 137

Slide 137 text

Chris J. Wright - [email protected] Empirical Study: Case Studies Case Study Tables Columns Primary Keys Foreign Keys Unique Constraints Cloc 2 10 0 0 0 JWhoisServer 6 49 6 0 0 NistDML182 2 32 1 1 0 NistDML183 2 6 0 1 1 RiskIt 13 56 11 10 0 UnixUsage 8 32 7 7 0

Slide 138

Slide 138 text

Chris J. Wright - [email protected] Empirical Study: Case Studies Case Study Tables Columns Primary Keys Foreign Keys Unique Constraints Cloc 2 10 0 0 0 JWhoisServer 6 49 6 0 0 NistDML182 2 32 1 1 0 NistDML183 2 6 0 1 1 RiskIt 13 56 11 10 0 UnixUsage 8 32 7 7 0

Slide 139

Slide 139 text

Chris J. Wright - [email protected] Empirical Study: Case Studies Case Study Tables Columns Primary Keys Foreign Keys Unique Constraints Cloc 2 10 0 0 0 JWhoisServer 6 49 6 0 0 NistDML182 2 32 1 1 0 NistDML183 2 6 0 1 1 RiskIt 13 56 11 10 0 UnixUsage 8 32 7 7 0

Slide 140

Slide 140 text

Chris J. Wright - [email protected] Empirical Study: Case Studies Case Study Tables Columns Primary Keys Foreign Keys Unique Constraints Cloc 2 10 0 0 0 JWhoisServer 6 49 6 0 0 NistDML182 2 32 1 1 0 NistDML183 2 6 0 1 1 RiskIt 13 56 11 10 0 UnixUsage 8 32 7 7 0

Slide 141

Slide 141 text

Chris J. Wright - [email protected] Empirical Study: Case Studies Case Study Tables Columns Primary Keys Foreign Keys Unique Constraints Cloc 2 10 0 0 0 JWhoisServer 6 49 6 0 0 NistDML182 2 32 1 1 0 NistDML183 2 6 0 1 1 RiskIt 13 56 11 10 0 UnixUsage 8 32 7 7 0

Slide 142

Slide 142 text

Chris J. Wright - [email protected] Empirical Study: Case Studies Case Study Tables Columns Primary Keys Foreign Keys Unique Constraints Cloc 2 10 0 0 0 JWhoisServer 6 49 6 0 0 NistDML182 2 32 1 1 0 NistDML183 2 6 0 1 1 RiskIt 13 56 11 10 0 UnixUsage 8 32 7 7 0

Slide 143

Slide 143 text

Chris J. Wright - [email protected] Empirical Study: Case Studies Case Study Total Constraints Total Mutants Cloc 0 30 JWhoisServer 50 184 NistDML182 2 66 NistDML183 2 18 RiskIt 36 160 UnixUsage 23 69

Slide 144

Slide 144 text

Chris J. Wright - [email protected] Empirical Study: DBMSs

Slide 145

Slide 145 text

Chris J. Wright - [email protected] Empirical Study: DBMSs PostgreSQL

Slide 146

Slide 146 text

Chris J. Wright - [email protected] Empirical Study: DBMSs SQLite PostgreSQL

Slide 147

Slide 147 text

Chris J. Wright - [email protected] Empirical Study: DBMSs SQLite PostgreSQL Client-Server Model

Slide 148

Slide 148 text

Chris J. Wright - [email protected] Empirical Study: DBMSs SQLite PostgreSQL Client-Server Model Local Client Model

Slide 149

Slide 149 text

Chris J. Wright - [email protected] Empirical Study: DBMSs SQLite PostgreSQL Client-Server Model Local Client Model Simultaneous Read/Write

Slide 150

Slide 150 text

Chris J. Wright - [email protected] Empirical Study: DBMSs SQLite PostgreSQL Client-Server Model Local Client Model Simultaneous Read/Write Locking on Write

Slide 151

Slide 151 text

Chris J. Wright - [email protected] Empirical Study: DBMSs SQLite PostgreSQL Client-Server Model Local Client Model Simultaneous Read/Write Locking on Write Prevents Parallel Approaches

Slide 152

Slide 152 text

Chris J. Wright - [email protected] Results

Slide 153

Slide 153 text

Chris J. Wright - [email protected] Results • Median of repetitions

Slide 154

Slide 154 text

Chris J. Wright - [email protected] Results • Median of repetitions • Lower-is-better metric

Slide 155

Slide 155 text

Chris J. Wright - [email protected] Results • Median of repetitions • Lower-is-better metric • Split by...

Slide 156

Slide 156 text

Chris J. Wright - [email protected] Results • Median of repetitions • Lower-is-better metric • Split by... ...case study

Slide 157

Slide 157 text

Chris J. Wright - [email protected] Results • Median of repetitions • Lower-is-better metric • Split by... ...case study ...DBMS

Slide 158

Slide 158 text

Chris J. Wright - [email protected] Results • Median of repetitions • Lower-is-better metric • Split by... ...case study ...DBMS • Full details in paper (including statistics)

Slide 159

Slide 159 text

Chris J. Wright - [email protected] 0 1000 2000 3000 Original Full Schemata Minimal Schemata Up−Front Schemata Just−in−Time Schemata Mutation Analysis Technique Mutation Analysis Time (ms) Original Full Schemata Minimal Schemata Up−Front Schemata Just−in−Time Schemata Postgres – Cloc Original Full Minimal Up-Front Just-in-Time

Slide 160

Slide 160 text

Chris J. Wright - [email protected] 0 1000 2000 3000 Original Full Schemata Minimal Schemata Up−Front Schemata Just−in−Time Schemata Mutation Analysis Technique Mutation Analysis Time (ms) Original Full Schemata Minimal Schemata Up−Front Schemata Just−in−Time Schemata Postgres – Cloc ~3.27s Original Full Minimal Up-Front Just-in-Time

Slide 161

Slide 161 text

Chris J. Wright - [email protected] 0 1000 2000 3000 Original Full Schemata Minimal Schemata Up−Front Schemata Just−in−Time Schemata Mutation Analysis Technique Mutation Analysis Time (ms) Original Full Schemata Minimal Schemata Up−Front Schemata Just−in−Time Schemata Postgres – Cloc ~3.27s ~2.08s Original Full Minimal Up-Front Just-in-Time

Slide 162

Slide 162 text

Chris J. Wright - [email protected] 0 50000 100000 150000 200000 250000 Original Full Schemata Minimal Schemata Up−Front Schemata Just−in−Time Schemata Mutation Analysis Technique Mutation Analysis Time (ms) Original Full Schemata Minimal Schemata Up−Front Schemata Just−in−Time Schemata Postgres – RiskIt Original Full Minimal Up-Front Just-in-Time

Slide 163

Slide 163 text

Chris J. Wright - [email protected] 0 50000 100000 150000 200000 250000 Original Full Schemata Minimal Schemata Up−Front Schemata Just−in−Time Schemata Mutation Analysis Technique Mutation Analysis Time (ms) Original Full Schemata Minimal Schemata Up−Front Schemata Just−in−Time Schemata Postgres – RiskIt ~238s Original Full Minimal Up-Front Just-in-Time

Slide 164

Slide 164 text

Chris J. Wright - [email protected] 0 50000 100000 150000 200000 250000 Original Full Schemata Minimal Schemata Up−Front Schemata Just−in−Time Schemata Mutation Analysis Technique Mutation Analysis Time (ms) Original Full Schemata Minimal Schemata Up−Front Schemata Just−in−Time Schemata Postgres – RiskIt ~238s ~225s Original Full Minimal Up-Front Just-in-Time

Slide 165

Slide 165 text

Chris J. Wright - [email protected] 0 50000 100000 150000 200000 250000 Original Full Schemata Minimal Schemata Up−Front Schemata Just−in−Time Schemata Mutation Analysis Technique Mutation Analysis Time (ms) Original Full Schemata Minimal Schemata Up−Front Schemata Just−in−Time Schemata Postgres – RiskIt ~238s ~225s ~23s Original Full Minimal Up-Front Just-in-Time

Slide 166

Slide 166 text

Chris J. Wright - [email protected] Results – Postgres

Slide 167

Slide 167 text

Chris J. Wright - [email protected] Results – Postgres ‘Full Schemata’

Slide 168

Slide 168 text

Chris J. Wright - [email protected] Results – Postgres ‘Full Schemata’ Improvement decreases with larger schemas

Slide 169

Slide 169 text

Chris J. Wright - [email protected] Results – Postgres ‘Minimal Schemata’ ‘Full Schemata’ Improvement decreases with larger schemas

Slide 170

Slide 170 text

Chris J. Wright - [email protected] Results – Postgres ‘Minimal Schemata’ ‘Full Schemata’ Improvement decreases with larger schemas Consistently faster, scales very well

Slide 171

Slide 171 text

Chris J. Wright - [email protected] Results – Postgres ‘Just-in-Time Schemata’ ‘Minimal Schemata’ ‘Full Schemata’ Improvement decreases with larger schemas Consistently faster, scales very well

Slide 172

Slide 172 text

Chris J. Wright - [email protected] Results – Postgres ‘Just-in-Time Schemata’ Consistently faster, scales very well ‘Minimal Schemata’ ‘Full Schemata’ Improvement decreases with larger schemas Consistently faster, scales very well

Slide 173

Slide 173 text

Chris J. Wright - [email protected] Results – Postgres ‘Up-Front Schemata’ ‘Just-in-Time Schemata’ Consistently faster, scales very well ‘Minimal Schemata’ ‘Full Schemata’ Improvement decreases with larger schemas Consistently faster, scales very well

Slide 174

Slide 174 text

Chris J. Wright - [email protected] Results – Postgres ‘Up-Front Schemata’ ‘Just-in-Time Schemata’ Consistently faster, scales very well Improvement decreases with larger schemas ‘Minimal Schemata’ ‘Full Schemata’ Improvement decreases with larger schemas Consistently faster, scales very well

Slide 175

Slide 175 text

Chris J. Wright - [email protected] Results – Postgres ‘Up-Front Schemata’ ‘Just-in-Time Schemata’ Consistently faster, scales very well Improvement decreases with larger schemas ‘Minimal Schemata’ ‘Full Schemata’ Improvement decreases with larger schemas Consistently faster, scales very well

Slide 176

Slide 176 text

Chris J. Wright - [email protected] 0 5000 10000 15000 Original Full Schemata Minimal Schemata Mutation Analysis Technique Mutation Analysis Time (ms) Original Full Schemata Minimal Schemata SQLite – Cloc Original Full Minimal

Slide 177

Slide 177 text

Chris J. Wright - [email protected] 0 5000 10000 15000 Original Full Schemata Minimal Schemata Mutation Analysis Technique Mutation Analysis Time (ms) Original Full Schemata Minimal Schemata SQLite – Cloc ~18.0s Original Full Minimal

Slide 178

Slide 178 text

Chris J. Wright - [email protected] 0 5000 10000 15000 Original Full Schemata Minimal Schemata Mutation Analysis Technique Mutation Analysis Time (ms) Original Full Schemata Minimal Schemata SQLite – Cloc ~18.0s ~18.7s Original Full Minimal

Slide 179

Slide 179 text

Chris J. Wright - [email protected] 0 5000 10000 15000 Original Full Schemata Minimal Schemata Mutation Analysis Technique Mutation Analysis Time (ms) Original Full Schemata Minimal Schemata SQLite – Cloc ~18.0s ~18.7s ~8.28s Original Full Minimal

Slide 180

Slide 180 text

Chris J. Wright - [email protected] 0 500000 1000000 1500000 2000000 Original Full Schemata Minimal Schemata Mutation Analysis Technique Mutation Analysis Time (ms) Original Full Schemata Minimal Schemata SQLite – RiskIt Original Full Minimal

Slide 181

Slide 181 text

Chris J. Wright - [email protected] 0 500000 1000000 1500000 2000000 Original Full Schemata Minimal Schemata Mutation Analysis Technique Mutation Analysis Time (ms) Original Full Schemata Minimal Schemata SQLite – RiskIt ~20.2 min Original Full Minimal

Slide 182

Slide 182 text

Chris J. Wright - [email protected] 0 500000 1000000 1500000 2000000 Original Full Schemata Minimal Schemata Mutation Analysis Technique Mutation Analysis Time (ms) Original Full Schemata Minimal Schemata SQLite – RiskIt ~20.2 min ~31.8 min Original Full Minimal

Slide 183

Slide 183 text

Chris J. Wright - [email protected] 0 500000 1000000 1500000 2000000 Original Full Schemata Minimal Schemata Mutation Analysis Technique Mutation Analysis Time (ms) Original Full Schemata Minimal Schemata SQLite – RiskIt ~20.2 min ~31.8 min ~2.5 min Original Full Minimal

Slide 184

Slide 184 text

Chris J. Wright - [email protected] Results – SQLite

Slide 185

Slide 185 text

Chris J. Wright - [email protected] Results – SQLite ‘Full Schemata’

Slide 186

Slide 186 text

Chris J. Wright - [email protected] Results – SQLite ‘Full Schemata’ Increasingly worsened with larger schemas

Slide 187

Slide 187 text

Chris J. Wright - [email protected] Results – SQLite ‘Minimal Schemata’ ‘Full Schemata’ Increasingly worsened with larger schemas

Slide 188

Slide 188 text

Chris J. Wright - [email protected] Results – SQLite ‘Minimal Schemata’ ‘Full Schemata’ Increasingly worsened with larger schemas Consistently faster, scales very well

Slide 189

Slide 189 text

Chris J. Wright - [email protected] Results – SQLite ‘Minimal Schemata’ ‘Full Schemata’ Increasingly worsened with larger schemas Consistently faster, scales very well

Slide 190

Slide 190 text

Chris J. Wright - [email protected] Future Work & Limitations

Slide 191

Slide 191 text

Chris J. Wright - [email protected] Future Work & Limitations Case Studies

Slide 192

Slide 192 text

Chris J. Wright - [email protected] Future Work & Limitations Case Studies DBMSs

Slide 193

Slide 193 text

Chris J. Wright - [email protected] Future Work & Limitations Detailed Timing Case Studies DBMSs

Slide 194

Slide 194 text

Chris J. Wright - [email protected] Future Work & Limitations Approaches Detailed Timing Case Studies DBMSs

Slide 195

Slide 195 text

Chris J. Wright - [email protected] Future Work & Limitations Approaches Detailed Timing Case Studies DBMSs Parallel Configurations

Slide 196

Slide 196 text

Chris J. Wright - [email protected] Future Work & Limitations Test Suite Approaches Detailed Timing Case Studies DBMSs Parallel Configurations

Slide 197

Slide 197 text

Chris J. Wright - [email protected] Test Suite Generation

Slide 198

Slide 198 text

Chris J. Wright - [email protected] Test Suite Generation • SchemaAnalyst tool

Slide 199

Slide 199 text

Chris J. Wright - [email protected] Test Suite Generation • SchemaAnalyst tool • Gregory Kapfhammer

Slide 200

Slide 200 text

Chris J. Wright - [email protected] Test Suite Generation • SchemaAnalyst tool • Gregory Kapfhammer • Tuesday 11:00am, Research & Industrial Track

Slide 201

Slide 201 text

Chris J. Wright - [email protected] Conclusions

Slide 202

Slide 202 text

Chris J. Wright - [email protected] Conclusions • Mutant Schemata and Parallelisation can both reduce the cost of mutation analysis

Slide 203

Slide 203 text

Chris J. Wright - [email protected] Conclusions • Mutant Schemata and Parallelisation can both reduce the cost of mutation analysis • The ‘Minimal Schemata’ approach...

Slide 204

Slide 204 text

Chris J. Wright - [email protected] Conclusions • Mutant Schemata and Parallelisation can both reduce the cost of mutation analysis • The ‘Minimal Schemata’ approach... ...consistently faster than original

Slide 205

Slide 205 text

Chris J. Wright - [email protected] Conclusions • Mutant Schemata and Parallelisation can both reduce the cost of mutation analysis • The ‘Minimal Schemata’ approach... ...consistently faster than original ...gives a reduction of up to 10x

Slide 206

Slide 206 text

Chris J. Wright - [email protected] Conclusions • Mutant Schemata and Parallelisation can both reduce the cost of mutation analysis • The ‘Minimal Schemata’ approach... ...consistently faster than original ...gives a reduction of up to 10x ...scales very well (for our case studies)

Slide 207

Slide 207 text

Chris J. Wright - [email protected] Conclusions • Mutant Schemata and Parallelisation can both reduce the cost of mutation analysis • The ‘Minimal Schemata’ approach... ...consistently faster than original ...gives a reduction of up to 10x ...scales very well (for our case studies) ...doesn’t require parallel DBMS access

Slide 208

Slide 208 text

Chris J. Wright - [email protected] Conclusions • Mutant Schemata and Parallelisation can both reduce the cost of mutation analysis • The ‘Minimal Schemata’ approach... ...consistently faster than original ...gives a reduction of up to 10x ...scales very well (for our case studies) ...doesn’t require parallel DBMS access • Website: http://schemaanalyst.org/

Slide 209

Slide 209 text

Chris J. Wright - [email protected] Conclusions • Mutant Schemata and Parallelisation can both reduce the cost of mutation analysis • The ‘Minimal Schemata’ approach... ...consistently faster than original ...gives a reduction of up to 10x ...scales very well (for our case studies) ...doesn’t require parallel DBMS access • Website: http://schemaanalyst.org/