dbflock
Run SQL scripts to upgrade and downgrade PostgreSQL database schemas.
How it works
dbflock works on a directory of sequentially-numbered integer schema versions starting from zero, with each version being a folder containing two scripts:
- apply.sql: script to execute when downgrading to the previous version
- revert.sql: script to execute when upgrading from the previous version
dbflock can then migrate to a version by applying scripts in order for all versions between the current and target.
For example, if the current database schema version is 12 and the target is 10, 12/revert.sql
and 11/revert.sql
will be applied in that order.
If the current database schema version is 12 and the target is 14, 13/apply.sql
and 14/apply.sql
will be applied in that order.
The current version, and a history of schema migrations, are stored in a table called dbflock_migration_history
, which is created automatically.
Usage
The standard libpq environment variables will be used to connect to the database.
Get the CLI
dbflock is a Node.js application available on npm, so Node.js is required. To install it globally as a command line application:
npm i -g dbflock
Make sure that npm's bin
directory is in the PATH
environment variable. The directory can be found using npm bin -g
.
Migrate to a schema version
The migration behaviour is described in the How it works section. If -v
is omitted, the highest schema version is used.
For full options, see dbflock migrate --help
.
dbflock migrate \
-s /path/to/schemas/ \
-v 42
Set the current specific schema version
This can be useful if the current database is already a specific schema version due to application manually or through other tools.
This does not apply any schema, and only updates the current schema version recorded in dbflock's internal table.
For full options, see dbflock set --help
.
dbflock set \
-v 42
Get the current specific schema version
For full options, see dbflock get --help
.
dbflock get