This week, I had an argument (a healthy one π) at work about the version in Rails' schema.rb.
In my feature branch, I added a migration file with version older than the latest version in main branch, so when I run db:migrate
, the version in schema.rb is not updated. My colleage insists that anytime you commit a new migration file, the schema.rb version needs to be updated too, or rails db:migrate
will not run that migration file in other environments. However, as I understand rails db:migrate
does not care about the version in schema.rb, it only compares the version list in db/migrate
folder and version list in schema_migrations
table, and run any migration that has not been run.
We didn't have much time to talk about it considering that it's not really important, so I just change the migration file name, run db:migrate
again in my local to update the version and commit it. Now, out of curiosity, I want to look more into this matter.
Rails' official guide
The first place to look is always the official guide.
By default, Rails generates db/schema.rb which attempts to capture the current state of your database schema.
It tends to be faster and less error prone to create a new instance of your application's database by loading the schema file via bin/rails db:schema:load than it is to replay the entire migration history. Old migrations may fail to apply correctly if those migrations use changing external dependencies or rely on application code which evolves separately from your migrations.
And here.
Note that running the db:migrate command also invokes the db:schema:dump command, which will update your db/schema.rb file to match the structure of your database.
It's just like what I see about schema.rb: just a file to store current table structure of your database. db:migrate
will generate or modify it, but it should not affect how db:migrate
run.
Stackoverflow
If answers and replies in this question are correct, it seems like db:migrate
used to assume that all migrations up to the version in schema.rb are applied, and will not run new migration files with older version. However, that has been already changed, and now db:migrate
does not decide which file should be run based on that version anymore.
Source code
Truth can be found in one place: the code.
γΌ Robert C. Martin, Clean Code
OK, enough for documents and answers. Now I need to see the truth. After searching through activerecord's source code, I confirmed my thoght: db:migrate
will check schema_migrations
table to determine which migration files have been run and which not. Then it will run all files of versions which are not in schema_migrations
def migrated
@migrated_versions || load_migrated
end
def load_migrated
@migrated_versions = Set.new(@schema_migration.all_versions.map(&:to_i))
end
def ran?(migration)
migrated.include?(migration.version.to_i)
end
def runnable
runnable = migrations[start..finish]
if up?
runnable.reject { |m| ran?(m) }
else
# skip the last migration if we're headed down, but not ALL the way down
runnable.pop if target
runnable.find_all { |m| ran?(m) }
end
end
Test
Of course, we cannot have a firm conclusion if we don't test it. So here is how I test it:
- Create a migration file with timestamp
20220129000000
in branchmain
and migrate it and commit schema.rb - Create a new branch
test
frommain
, add a migration file with timestamp20220128000000
and migrate it => schema.rb version is expected not to change - Change back to
main
branch, drop db and migrate again - Merge
test
intomain
and run migrate again => although schema.rb version does not change, the migration file20220128000000
is expected to be applied
Result
After step 2, schema.rb version does not change:
After step 3, only the migration file of version 20220129000000
After step 4, the migration file of verion 20220128000000
is applied even when schema.rb version does not change:
Phew... Finally I got something to show my colleage the next time we talk! π§
(Post copied from my blog)
Top comments (0)