Name
sqitch - Sensible database change management
Synopsis
sqitch <command> [options] [command-options] [args]
Description
Sqitch is a database change management application. It currently supports:
- PostgreSQL 8.4+
- YugabyteDB 2.6+
- CockroachDB 21+
- SQLite 3.8.6+
- MySQL 5.1+
- MariaDB 10.0+
- Oracle 10g+,
- Firebird 2.0+
- Vertica 7.2+
- Exasol 6.0+
- Snowflake
What makes it different from your typical migration-style approaches? A few things:
-
No opinions
Sqitch is not tied to any framework, ORM, or platform. Rather, it is a standalone change management system with no opinions about your database engine, application framework, or development environment.
-
Native scripting
Changes are implemented as scripts native to your selected database engine. Writing a PostgreSQL application? Write SQL scripts for
psql
. Writing an Oracle-backed app? Write SQL scripts for SQL*Plus. -
Dependency resolution
Database changes may declare dependencies on other changes – even on changes from other Sqitch projects. This ensures proper order of execution, even when you’ve committed changes to your VCS out-of-order.
-
Deployment integrity
Sqitch manages changes and dependencies via a plan file, and employs a Merkle tree pattern similar to Git and Blockchain to ensure deployment integrity. As such, there is no need to number your changes, although you can if you want. Sqitch doesn’t much care how you name your changes.
-
Iterative Development
Up until you tag and release your project, you can modify your change deployment scripts as often as you like. They’re not locked in just because they’ve been committed to your VCS. This allows you to take an iterative approach to developing your database schema. Or, better, you can do test-driven database development.
Ready to get started? Here’s where:
-
Sqitch Tutorials
Detailed tutorials demonstrating the creation, development, and maintenance of a database with Sqitch.
-
Slides from “Sane Database Management with Sqitch”, presented to the Portland Perl Mongers in January, 2013.
-
Movie of “Sane Database Management with Sqitch”, presented to the Portland PostgreSQL Users Group in September, 2012.
-
Three-hour tutorial session on using Git, test-driven development with pgTAP, and change management with Sqitch.
Terminology
-
change
A named unit of change. A change name must be used in the file names of its deploy and a revert scripts. It may also be used in a verify script file name.
-
tag
A known deployment state, pointing to a single change, typically corresponding to a release. Think of it is a version number or VCS revision. A given point in the plan may have any number of tags.
-
state
The current state of the database. This is represented by the most recently-deployed change. If the state of the database is the same as the most recent change, then it is considered “up-to-date”.
-
plan
A list of one or more changes and their dependencies that define the order of deployment execution. The plan is stored in a “plan file,” usually named
sqitch.plan
. Sqitch reads the plan file to determine what changes to execute to change the database from one state to another. -
target
A named database to which to deploy changes. Always has an associated connection URI, and may also have an associated command-line client and registry name.
-
registry
The name of the database object where Sqitch’s state and history data is stored. Typically a schema name (as in PostgreSQL and Oracle) or a database name (as in SQLite and MySQL).
-
add
The act of adding a change to the plan. Sqitch will generate scripts for the change, which you then may modify with the necessary code (typically DDLs) to actually deploy, revert, and verify the change.
-
deploy
The act of deploying changes to a database. Sqitch reads the plan, checks the current state of the database, and applies all the changes necessary to either bring the database up-to-date or to a requested state (a change name or tag).
-
revert
The act of reverting database changes to reach an earlier deployment state. Sqitch reads the list of deployed changes from the database and reverts them in the reverse of the order in which they were applied. All changes may be reverted, or changes may be reverted to a requested state (a change name or tag).
-
committer
User who commits or reverts changes to a database.
-
planner
User who adds a change to the plan.
Options
-C --chdir --cd DIR Change to directory before performing any actions.
--etc-path Print path to etc directory and exit.
--no-pager Do not pipe output into a pager.
--quiet Quiet mode with non-error output suppressed.
-V --verbose Increment verbosity.
--version Print version number and exit.
--help Show a list of commands and exit.
--man Print introductory documentation and exit.
Options Details
-
--chdir
-
--cd
-
-C
sqitch --chdir dbproject sqitch --cd /usr/local/somedb sqitch -C dbcheckout
Change to the specified directory before performing any actions. Effectively the same as:
(cd somedir && sqitch ...)
But a bit friendlier when managing multiple projects.
-
--etc-path
sqitch --etc-path
Print out the path to the Sqitch
etc
directory and exit. This is the directory where the system-wide configuration file lives, as well as change script templates. -
--no-pager
sqitch --no-pager
Do not pipe Sqitch output into a pager. Currently limited to the
log
andplan
commands. -
--quiet
sqitch --quiet
Suppress normal output messages. Error messages will still be emitted to
STDERR
. Overrides any value specified by--verbose
. -
-V
-
--verbose
sqitch --verbose sqitch -VVV
Pass multiple times to specify a value between 0 and 3 to determine how verbose Sqitch should be. Unless
--quiet
is specified, the default is 1, meaning that Sqitch will output basic status messages as it does its thing. Values of 2 and 3 each cause greater verbosity. Ignored if--quiet
is specified. -
--help
sqitch --help
Outputs a brief description all known Sqitch commands and exits.
-
--man
sqitch --man
Outputs this documentation and exits.
-
--version
sqitch --version
Outputs the program name and version and exits.
Sqitch Commands
-
Create the plan file and directories for deploy, revert, and verify scripts if they do not already exist. This command is useful for starting a new Sqitch project.
-
Output information about the current deployment state of a database, including the name of the last deployed change, as well as any tags applied to it. If any changes in the plan have not been deployed, they will be listed separately.
-
Search and Output the complete change history of a database. Provides information about when changes were deployed, reverted, or failed, as well as who planned and committed the changes, and when.
-
Add a new change.
-
List tags or tag the latest change.
-
Rework an existing change.
-
Manage target databases.
-
Deploy changes to a database
-
Revert changes from a database.
-
Verify changes deployed to a database.
-
Get and set project, user, or system Sqitch options.
-
Bundle a Sqitch project for distribution. This command copies the Sqitch configuration, plan, and deploy, revert, and verify scripts to a directory, so that it can be packaged up for distribution, such as in an RPM or tarball.
-
Show help for a specific command or, if no command is specified, show the same documentation as
--help
.
Configuration
Sqitch configuration can be set up on a project, user, or system-wide basis.
The format of the configuration file, named sqitch.conf
, is the same as for
git.
Here’s an example of a configuration file that might be useful checked into a
VCS for a project that deploys to PostgreSQL and stores its deployment scripts
with the extension ddl
under the migrations
directory. It also wants
bundle to be created in the _build/sql
directory, and to deploy starting
with the “gamma” tag:
[core]
engine = pg
top_dir = migrations
extension = ddl
[engine "pg"]
target = widgetopolis
[revert]
to = gamma
[bundle]
from = gamma
tags_only = yes
dest_dir = _build/sql
[target "widgetopolis"]
uri = db:pg:widgetopolis
And here’s an example of useful configuration in ~/.sqitch/sqitch.conf
, to
point to system-specific engine information:
[user]
name = Marge N. O’Vera
email = marge@example.com
[engine "pg"]
client = /usr/local/pgsql/bin/psql
[engine "mysql"]
client = /usr/local/mysql/bin/mysql
[engine "oracle"]
client = /usr/local/instantclient_11_2/sqlplus
[engine "sqlite"]
client = /usr/local/bin/sqlite3
Various commands read from the configuration file and adjust their operation accordingly. See sqitch-config for a list.
See Also
The original design for Sqitch was sketched out in a number of blog posts:
- Simple SQL Change Management
- VCS-Enabled SQL Change Management
- SQL Change Management Sans Duplication
Other tools that do database change management include:
-
Numbered migrations for Ruby on Rails.
-
Numbered changes in pure SQL, integrated with Perl’s Module::Build build system. Does not support reversion.
-
Numbered migrations in pure SQL.
-
PostgreSQL-specific dependency-tracking solution by depesz.
Author
David E. Wheeler david@justatheory.com
License
Copyright (c) 2012-2024 iovation Inc., David E. Wheeler
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the “Software”), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.