Offline Database Support
If you cannot run Liquibase directly against a database, there are two major options to ensure your database is kept up to date.
The most common way to manage “offline” databases is to use the
updateSQL command reads a database's DATABASECHANGELOG table and outputs the SQL that would run as part of the upgrade.
This SQL includes inserts into the DATABASECHANGELOG table and can be ran against any database to both upgrade it and keep its history correct.
It is important that the database you generate the SQL from is the same as the database(s) you plan to run the SQL against.
Unless you have preconditions in your changelog file, the
updateSQL process only reads the DATABASECHANGELOG table to determine what changesets to run.
Therefore if, for example, you have a production database that you cannot run Liquibase against directly and it may be different than your test databases,
you can copy/restore just the DATABASECHANGELOG table from the production database into a database you can run
updateSQL against. Then run the generated SQL against the actual production database.
If your environment does not lend itself to using
updateSQLwith test and backup databases, you can run Liquibase using an “offline” database.
Running in offline mode only supports
tagExists. It does not support direct
diff, or preconditions as there is nothing to actually update or state to check.
An offline database is “connected” to using a URL syntax of
|changeLogFile||Specify the file acting as the DATABASECHANGELOG table. Defaults to
|version||Specify the database version to ensure generated SQL matches target database version. Example: 5.4.2 or 18.104.22.168|
|productName||Specify “product name” seen by the JDBC driver.|
|catalog||Specify the connection catalog|
|caseSensitive||Specify if the database is case sensitive or not|
|outputLiquibaseSql||If set to “true”, output from