Create a database patch¶

Launchpad’s schema is changed through database patches. These live in branches that go through review and landing, just like code changes.

Schema patches can either be applied while the system has deliberately been brought down (cold patches) or while the system is up and actively being used (hot patches). Changes such as updating a table, which will be affected by ongoing activity, and adding an index, which can be done without affecting behavior of the system, must be split into two separate patches — one cold and one hot.

See Database patching for more information about how cold/hot patching works.

Schema patches must not be combined with changes to the Launchpad python code — any branch landing on the devel or db-devel must either be a schema patch or be a code change. Test only code may be included if absolutely necessary. Exceptions to this rule require approval from the project lead, because deploying them will require a 1 hour+ complete downtime window.

Making a database patch¶

You need to run these steps whenever you make a schema change:

In your local development Launchpad instance, run make schema to get a pristine database of sample data.
Add a patch number in the dbpatches repository. If you are in the ~launchpad team please allocate this yourself. Other developers can ask any ~launchpad member to allocate a patch number for them. Commit and push directly to the main branch. No review is needed.
Create a SQL file named patch-XXXX-YY-Z.sql in database/schema/, where XXXX-YY-Z is the patch number allocated in the previous step. The schema application code only picks up files matching this exact naming pattern.

Your patch should look like this:

-- Copyright 2026 Canonical Ltd.  This software is licensed under the
-- GNU Affero General Public License version 3 (see the file LICENSE).

SET client_min_messages=ERROR;

-- put your changes in here.

INSERT INTO LaunchpadDatabaseRevision VALUES (XXXX, YY, Z);

Run your new SQL patch on the development database to ensure that it works:
```
psql launchpad_dev -1 -f your-patch.sql
```
Run make schema again and check that your patch gets applied. This is needed to allow the test suite to see your changes.
You may also wish to run make newsampledata. Although it isn’t critical, this lets you see what changes your patch would make to initial setups. This will produce a lot of noise. Feel free to ignore it.
Review the sample data changes that occurred using git diff database/sampledata. This diff can be hard to review as-is. You might want to use a graphical diff viewer like kompare or meld. Make sure that you understand all the changes you see.
If you have added, removed or renamed a table or column, ensure that your patch includes appropriate COMMENT statements.
Run make schema again to ensure that it works, and that you now have a pristine database with the new sample data. If you don’t want to clean your launchpad_dev database, then you can use make -C database/schema test instead to update only the test databases.
Make any necessary changes to database/schema/fti.py, database/schema/security.cfg.
Run the full test suite to ensure that your new schema doesn’t accidentally break any existing tests/code.

Propose your patch for deployment

Commit your patch without sample data changes, then push and propose for merging to db-devel if it’s a cold patch or master if it’s a hot patch.

The schema change is approved for landing when you have an ‘Approved’ vote from a DB reviewer (unless the reviewer in question explicitly sets a higher barrier).

In case you’re not a member of ~launchpad, ask a member of the team to do these last steps for you:

QA the patch to confirm it applies cleanly and meets performance requirements. Make sure cold patches do not take more than a few seconds to apply.
Wait until there are no blockers to deploying the patch. One common blocker is having code changes made prior to the patch sitting in stable and not yet deployed to all affected service instances.
Request a deployment per the internal production change process.

Rules for patches¶

Operations that claim locks on tables should be done using CONCURRENTLY.
When dropping a table, make sure that you drop or update any dependent triggers, views and foreign keys beforehand.
Do not migrate data in schema patches unless the data size is extraordinarily small (< 100s of rows).
Similarly, new columns must default NULL unless the data size is extraordinarily small (< 100s of rows).
When changing existing DB functions, start your patch with the original version (SELECT pg_get_functiondef(oid) FROM pg_proc WHERE proname IN ('foofunc', 'barfunc') ORDER BY proname;). This makes it much easier to review the diff.

Notes on changing security.cfg¶

If your changes to security.cfg are additive in nature like adding new permissions to an existing role or adding a new role, then you can land them on master rather than db-devel. These changes are deployed during no-downtime rollouts.

Any update that revokes permissions from an existing role (including removing a role entirely), must be landed on db-devel and deployed during a fast-downtime deployment. This is because a full update (including revocation) requires resetting all permissions, which cannot be done without downtime.

Adding new roles requires manual DB reconfiguration, so you need to file a ticket to grant access to relevant machines and make sure it is resolved before landing the branch that needs them.

Sample data¶

If your branch needs to make changes to the database schema, the sample data should be updated to match your schema changes.

Never add new rows to the sample data.

Sample data is used to provide well-known baseline data for the test suite, and to populate a developer’s Launchpad instance so that launchpad.test can display interesting stuff. Keep the following guidelines and recommendations in mind before you make changes to the test suite sample data, or you may break the tests for yourself or others.

Sample data is for developer’s instances only and is of no use on production systems.

If your tests require new data, create the data in your test’s harness instead of adding new sample data. This often makes the tests themselves more readable. It also reduces the chance that your changes will break other tests. Add the new data in your test’s setUp() or in the narrative of your doctest. The test suite uses the launchpad_ftest_template database, so there is no chance that running the test suite will accidentally alter the sample data.

However, if you interact with the web UI for launchpad.dev your changes will end up in the launchpad_dev database. This database is used to create the new sample data, so you must run make schema to start with a pristine database before generating new sample data. If in fact you do want the effects of your UI interactions to land in the new sample data, then the general process is to

run make schema
interact with launchpad.dev
run make newsampledata

Note

Be aware that if you generate new sample data, this will probably have an effect on tests not related to your changes. If you make changes to the sample data, you must run the full test suite and ensure that you get no failures.