SE2016 Java Roman Ugolnikov "Migration and source control for your DB"
0 likes137 views
The document discusses database migration and source control. It describes how database structure, data, and logic can change across versions. It recommends using tools like Liquibase and Flyway to manage database schema changes and keep the database schema in sync with code. These tools allow defining changes in SQL-based changesets and tracking them in version control. The document also covers how the tools implement features like rollback of changes, preconditions for changes, and support for multiple database types.
SE2016 Java Roman Ugolnikov "Migration and source control for your DB"
- 1. Migrationа and source control for your DB
- 2. what: Database migration: Database structure: - Tables, constrains, indexes; Database data: - Initial data like list of post codes, statuses for order, etc. Database logic: - stored procedures, triggers, functions
- 3. why: versions Data Base migration: Data base structure: - Tables, constrains, indexes; Data base data: - Initial data like list of post codes, statuses for order, etc. Data base logic: - stored procedures, triggers, functions v1.0.0
- 4. why: versions Data Base migration: Data base structure: - Tables, constrains, indexes; Data base data: v1.0.0
- 5. why: versions Data Base migration: Data base structure: - Tables, constrains, indexes; Data base data: - Initial data like list of post co v1.2.1 v1.0.0
- 6. Data Base migration: Data base structure: - Tables, constrains, indexes; Data base data: - Initial data like list of post v1.2.1 v1.6.0 v1.0.0 why: versions
- 7. Data Base migration: Data base structure: - Tables, constrains, indexes; Data base data: - Initial data like list of post c v1.2.1 v1.6.0 v2.0.1 v1.0.0 why: versions
- 8. Data Base migration: Data base structure: - Tables, constrains, indexes; Data base data: - Initial data like list of post codes, statuses for order, etc. Data base logic: - stored procedures, triggers, functions v1.2.1 v1.6.0 v2.0.1 v1.0.0 why: versions
- 9. Data Base migration: Data base structure: - Tables, constrains, indexes; Data base data: - Initial data like list of post codes, statuses for order, etc. Data base logic: - stored procedures, triggers, functions v1.2.1 v1.6.0 v2.0.1 v1.0.0DEV why: environments
- 10. Conclusion We need version control over our database migration
- 15. Tools Liquibase : •Apache license •Started in 2006 (active) •http://liquibase.org Flyway: •Apache license •Started in 2010 •http://www.flywaydb.or g
- 16. Common flow Build Test Run • Dev • Test • Acceptance • Production • GUI tests • Integration tests • API tests • Unit tests • Ant • Maven • Gradle
- 17. Common flow Build Test Run • Dev • Test • Acceptance • Production • GUI tests • Integration tests • API tests • Unit tests • Ant • Maven • Gradle
- 18. Common flow Build Test Run • Java API • Spring Bean • Plugins • CLI • Plugins • CLI
- 21. How it works Sequential change flow 1. Init db 2. Add column “type” 3. Add new table
- 22. How it works Sequential change flow 1. Init db 2. Add column “type” 3. Add new table X
- 23. How it works Sequential change flow 1. Init db 3. Add new table 4. Remove column “type” 2. Add column “type”
- 24. difference?, comparison. Rollback vs no rollback SQL output Conditions baseline Supported modes: https://flywaydb.org/documentation/gradle/ Flyway more and esear.
- 26. How it works Changes are grouped into changesets: • Change(s) that should be applied atomically Changesets are grouped into changelogs: •Files managed in version control
- 28. Supported db list MySQL PostgreSQL Oracle MSSQL Server Sybase_Enterprise Sybase_Anywhere DB2 Apache_Derby HSQL H2 Informix Firebird SQLite MySQL PostgreSQL Oracle MSSQL Server SQL Azure MariaDB Phoenix Vertica AWS Redshift DB2, DB2 z/OS Derby H2, Hsql, SQLite, SAP HANA, solidDB, Sybase ASE
- 29. Supported formats DSL for database changes: • XML • YAML • JSON • sql Pure SQL
- 31. Toolbox • SQL output • Rollback • Context • Precondition
- 32. Toolbox: precondition <preConditions onFail="WARN" onError="MARK_RAN"> <dbms type="oracle"/> <runningAs username="SYSTEM"/> <tableExists tableName=“newTable"/> <customPrecondition className=“com.example.CheckOne"/> <sqlCheck expectedResult="0"> select count(*) from oldtable </sqlCheck> </preConditions>
- 33. Toolbox: precondition <preConditions onFail="WARN" onError="MARK_RAN"> <dbms type="oracle"/> <runningAs username="SYSTEM"/> <tableExists tableName=“newTable"/> <customPrecondition className=“com.example.CheckOne"/> <sqlCheck expectedResult="0"> select count(*) from oldtable </sqlCheck> </preConditions>
- 34. Toolbox: precondition <preConditions onFail="WARN" onError="MARK_RAN"> <dbms type="oracle"/> <runningAs username="SYSTEM"/> <tableExists tableName=“newTable"/> <customPrecondition className=“com.example.CheckOne"/> <sqlCheck expectedResult="0"> select count(*) from oldtable </sqlCheck> </preConditions>
- 35. Toolbox: precondition <preConditions onFail="WARN" onError="MARK_RAN"> <dbms type="oracle"/> <runningAs username="SYSTEM"/> <tableExists tableName=“newTable"/> <customPrecondition className=“com.example.CheckOne"/> <sqlCheck expectedResult="0"> select count(*) from oldtable </sqlCheck> </preConditions>
- 36. Toolbox: precondition <preConditions onFail="WARN" onError="MARK_RAN"> <dbms type="oracle"/> <runningAs username="SYSTEM"/> <tableExists tableName=“newTable"/> <customPrecondition className=“com.example.CheckOne"/> <sqlCheck expectedResult="0"> select count(*) from oldtable </sqlCheck> </preConditions>
- 37. Toolbox: precondition <preConditions onFail="WARN" onError="MARK_RAN"> <dbms type="oracle"/> <runningAs username="SYSTEM"/> <tableExists tableName=“newTable"/> <customPrecondition className=“com.example.CheckOne"/> <sqlCheck expectedResult="0"> select count(*) from oldtable </sqlCheck> </preConditions>
- 38. Toolbox: precondition <preConditions onFail="WARN" onError="MARK_RAN"> <dbms type="oracle"/> <runningAs username="SYSTEM"/> <tableExists tableName=“newTable"/> <customPrecondition className=“com.example.CheckOne"/> <sqlCheck expectedResult="0"> select count(*) from oldtable </sqlCheck> </preConditions>
- 39. DEMO
- 40. let us try it DDLv1
- 41. DDL v2 v1 Init data, new column (source) let us try it
- 42. how it works DDL v2 v3 v1 Init data, new column (source) Move values in separate table let us try it
- 43. let us try it DDL v2 v3 v1 Init data, new column (source) Move values in separate table
- 44. Thank you for attention Q & A uholnikovroman@gmail.com