The document discusses the importance of database migration and the tools used, specifically Liquibase, for managing database schema changes and initial data. It covers the structure of database changes, how Liquibase works with various database types, and the formats supported for changesets and changelogs. Additionally, it outlines recommendations and peculiarities associated with using Liquibase for effective database version control.

1. source control for your DB

2. why
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

3. why
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
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
v1.0.0

5. why
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
v1.0.0

6. why
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

7. why
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

8. What
Liquibase :
•Apache license
•Started in 2006 (active)

9. What
Liquibase :
•Database migration for Java;
•Can be used as
•Ant, Maven or Gradle plugin,
•as CLI tool;
• as part of the system :
•Servlet Listener
•Spring Listener
•JEE CDI Listener

10. how it works
•Supports multiple database types:
MySQL, PostgreSQL, Oracle, MsSql,
Sybase_Enterprise, Sybase_Anywhere , DB2,
Apache_Derby, HSQL, H2, Informix, Firebird,
SQLite

11. 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

12. how it works
Supports XML, YAML, JSON (DSL for database changes)
and SQL formats:
•Create Table, Add PK, Add FK, Add
Column, Add Index, …
•Drop Table, Drop PK, Drop FK, Drop
Column, Drop Index, …
•Insert, Update, Delete, …

13. how it works
•Changeslog (XML):
<databaseChangeLog xmlns=…>
<changeSet author="liquibase-docs“
id="addColumn-example">
<addColumn catalogName="cat“
schemaName="public"
tableName="person">
<column name="address"
type="varchar(255)"/> </addColumn>
</changeSet>
<changeSet> ….</changeset>
<changeSet> ….</changeset>
</databaseChangeLog>

14. how it works
•Changeslog (YAML):
databaseChangeLog:
changeSet:
id: addColumn-example
author: liquibase-docs
changes:
- addColumn:
catalogName: cat
columns:
- column:
name: address
type: varchar(255)
schemaName: public
tableName: person

15. how it works
•Changeslog (JSON){
databaseChangeLog: [
"changeSet": {
"id": "addColumn-example",
"author": "liquibase-docs",
"changes": [
{
"addColumn": {
"catalogName": "cat",
"columns": [
{
"column": {
"name": "address",
"type": "varchar(255)“
}
}]
,
"schemaName": "public",
"tableName": "person"
}
}]
}

16. how it works
- Changesets uniquely identified by [Author, ID, File path]
- Liquibase tracks changeset execution in a special table
- Lock table to prevent concurrent Liquibase invocations
- Modified changesets are detected via checksums

17. how it works
DatabaseChangelogLock:
DatabaseChangelog:

18. how it works
CLI commands:
update [Count, Tag]
rollback [Count, Data, Tag]
generateChangeLog --diffTypes=tables, views,
columns, indexes, foreignkeys, primarykeys,
uniqueconstraints, data
diff --referenceUrl=<value>
status --verbose
updateSQL ...

19. let us try it
DDLv1

20. DDL
v2
v1
Init data,
new column
(source)
let us try it

21. how it works
DDL
v2
v3
v1
Init data,
new column
(source)
Move values in
separate table
let us try it

22. let us try it
DDL
v2
v3
v1
Init data,
new column
(source)
Move values in
separate table

23. how it works
DEMO

24. how it works
Spring config (embedded in your
application):
<dependency>
<groupId>org.liquibase</groupId>
<artifactId>liquibase-core</artifactId>
<version>${liquibase.version}</version>
</dependency>
<bean id="liquibase" class="liquibase.integration.spring.SpringLiquibase">
<property name="dataSource" ref="myDataSource" />
<property name="changeLog" value="classpath:db-changelog.xml" />
<!-- contexts specifies the runtime contexts to use. -->
<property name="contexts" value="test, production" />
</bean>

25. how it works
Spring Boot config (embedded in your
application ):
<dependency>
<groupId>org.liquibase</groupId>
<artifactId>liquibase-core</artifactId>
<version>${liquibase.version}</version>
</dependency>
none
https://github.com/spring-projects/spring-boot/tree/master/spring-boot-samples/spring-boot-sample-liquibase

26. how it works
•Also:
<changeSet id="4" dbms="oracle">
<sqlFile path="5.sql"/>
</changeSet>
<changeSet id="5" context="test">
<sqlFile path="5.sql"/>
</changeSet>
<changeSet id="6" failOnError="false">
<sqlFile path="6.sql"/>
</changeSet>

27. how it works
•Also:
<preConditions>
<dbms type="oracle" />
<runningAs username="SYSTEM" />
</preConditions>
<changeSet id="1" author="bob">
<preConditions onFail="WARN">
<sqlCheck expectedResult="0">
select count(*) from oldtable
</sqlCheck>
</preConditions>
<comment>Comments should go after preCondition. If they are before then
liquibase usually gives error.</comment>
<dropTable tableName="oldtable"/>
</changeSet>

28. recommendations
• it is better to have each change in separate changes ;
• Document changesets with <comment> tag;
• Folders structure should reflect migration structure
com
example
db
changelog
db.changelog-master.xml
db.changelog-1.0.xml
db.changelog-1.1.xml
db.changelog-2.0.xml

29. peculiarities
•if you do roll back, Liquibase deletes the log line about the update from
databasechangelog table, and after that Liguibase does not know any more
that this change was applied and reverted;
•all changes are applied sequentially, so we can not apply/rollback
separate mediate changes;
•no straightforward way of customizing warn/error messages;

30. peculiarities
•Each change set has an “id” and “author” attribute which, along with the directory
and file name of the the change log file, uniquely identifies it. Dark side of it is that
running the same change set using relative and absolute path will be considered
as different changesets!
For example we have file /home/user/changelog-0.1.0.xml
First we execute:
$ liquibase --changeLogFile=/home/user/changelog-0.1.0.xml
and then if we try to execute
$ liquibase --changeLogFile=changelog-0.1.0.xml

31. The end
Thank you for attention
Q & A
r.uholnikov@levi9.com