Custom table names

[gap777]

As of https://github.com/bensheldon/good_job/releases/tag/v2.11.2, good_jobs supported custom table names. However, this seems not to work anymore. Specifically, trying to set them in a config initializer

GoodJob::Process.table_name = 'my_good_job_processes'
GoodJob::ActiveJobJob.table_name = 'my_good_jobs'
GoodJob::Execution.table_name = 'my_good_jobs'

generates errors like:

.../config/initializers/good_job.rb:1:in `<main>': uninitialized constant GoodJob::Process (NameError)
17:06:22 web.1  |
17:06:22 web.1  | GoodJob::Process.table_name = 'my_good_job_processes'
17:06:22 web.1  |        ^^^^^^^^^

Is there another way to configure this, or is this capability lost?
Can it be restored?
Thank you!

[bensheldon]

@gap777 Thanks for opening the issue and sorry about the inconvenience!

Those models became autoloaded. You should be able to do this in an an initializer:

# config/initializers/good_job.rb

Rails.application.reloader.to_prepare do
  GoodJob::Execution.table_name = 'my_good_jobs'
  GoodJob::Process.table_name = 'my_good_job_processes'
  GoodJob::Setting.table_name = 'my_good_job_settings' # <- new database backed model
end

Edit: GoodJob::Job's table name does not need to be set directly.

[remy727]

@bensheldon, I am getting the following error with the above configuration.

/Users/admin/.rvm/gems/ruby-3.2.0/gems/good_job-3.8.0/app/models/good_job/job.rb:24:in `table_name=': Assign GoodJob::Execution.table_name directly (NotImplementedError)

What is the correct configuration in the latest version?

[bensheldon]

@remy727 oops, you don't need to set the table name for GoodJob::Job; it will automatically be set based on what GoodJob::Execution is named. I'll edit my original suggestion.

good_job/app/models/good_job/job.rb

Lines 21 to 25 in 430a6a0

	delegate :table_name, to: GoodJob::Execution
	def table_name=(_value)
	raise NotImplementedError, 'Assign GoodJob::Execution.table_name directly'
	end

[remy727]

Thank you for your reply @bensheldon
The error went way but I am still seeing the good_jobs not my_good_jobs in migration file.

[bensheldon]

@remy727 oh! Migrations/Schema and Model configuration aren't connected. If you want to rename the tables themselves, you'll have to write your own migrations.

[remy727]

Got it. Thank you for confirming @bensheldon!

[mokolabs]

I tried using the custom initializer code above and updated the migration to match my new table names.

But, when I relaunch the app after migrating, I'm seeing this error a lot in my logs:

DEPRECATION WARNING: GoodJob has pending database migrations. To create the migration files, run:
  rails generate good_job:update
To apply the migration files, run:
  rails db:migrate

Is there any way to disable that?

[bensheldon]

@mokolabs sorry about that! I maybe hardcoded a tablename in one of those checks 😰 I'll take a look because that is annoying.

[bensheldon]

I tried customizing them myself. I'm not seeing the migration warning, but I had to add a bunch more overrides:

Rails.application.reloader.to_prepare do
  GoodJob::BaseExecution.table_name = 'c_good_jobs'
  GoodJob::BatchRecord.table_name = 'c_good_jobs'
  GoodJob::DiscreteExecution.table_name = 'c_good_job_executions'
  GoodJob::Execution.table_name = 'c_good_jobs'
  GoodJob::Process.table_name = 'c_good_job_processes'
  GoodJob::Setting.table_name = 'c_good_job_settings'
end

[gap777]

That's what mine looks like these days:

# MONKEYPATCH
Rails.application.reloader.to_prepare do
  GoodJob::BaseExecution.table_name = 'core_jobs__jobs'
  GoodJob::Execution.table_name = 'core_jobs__jobs'
  GoodJob::DiscreteExecution.table_name = 'core_jobs__executions'
  GoodJob::Process.table_name = 'core_jobs__processes'
  GoodJob::Setting.table_name = 'core_jobs__settings'
  GoodJob::BatchRecord.table_name = 'core_jobs__batches'
end

[gap777]

FWIW, I'm not seeing the warning with the latest build... I think there one or two regressions in that particular area since the bug was originally fixed, though those have already been fixed as well in the latest.

[mokolabs]

@bensheldon @gap777 Thank you for the quick reply! I'll give it another go. 👍🏻

[mokolabs]

@bensheldon @gap777 That worked. Thanks again!

For those who wish to go down this path, here's everything you need to do it:

1. Create a good_job initializer at config/initializers/good_job.rb.
2. Add your custom table names to the initializer:

   # Rename good_job tables
   Rails.application.reloader.to_prepare do
     GoodJob::BaseExecution.table_name = 'jobs'
     GoodJob::BatchRecord.table_name = 'jobs_batches'
     GoodJob::DiscreteExecution.table_name = 'jobs_executions'
     GoodJob::Execution.table_name = 'jobs'
     GoodJob::Process.table_name = 'jobs_processes'
     GoodJob::Setting.table_name = 'jobs_settings'
   end
   

3. Tweak the default good_job database migration to use your custom table names:

   class CreateGoodJobs < ActiveRecord::Migration[7.0]
     def change
       # Uncomment for Postgres v12 or earlier to enable gen_random_uuid() support
       # enable_extension 'pgcrypto'
   
       create_table :jobs, id: :uuid do |t|
         t.text :queue_name
         t.integer :priority
         t.jsonb :serialized_params
         t.datetime :scheduled_at
         t.datetime :performed_at
         t.datetime :finished_at
         t.text :error
   
         t.timestamps
   
         t.uuid :active_job_id
         t.text :concurrency_key
         t.text :cron_key
         t.uuid :retried_good_job_id
         t.datetime :cron_at
   
         t.uuid :batch_id
         t.uuid :batch_callback_id
   
         t.boolean :is_discrete
         t.integer :executions_count
         t.text :job_class
         t.integer :error_event, limit: 2
       end
   
       create_table :jobs_batches, id: :uuid do |t|
         t.timestamps
         t.text :description
         t.jsonb :serialized_properties
         t.text :on_finish
         t.text :on_success
         t.text :on_discard
         t.text :callback_queue_name
         t.integer :callback_priority
         t.datetime :enqueued_at
         t.datetime :discarded_at
         t.datetime :finished_at
       end
   
       create_table :jobs_executions, id: :uuid do |t|
         t.timestamps
   
         t.uuid :active_job_id, null: false
         t.text :job_class
         t.text :queue_name
         t.jsonb :serialized_params
         t.datetime :scheduled_at
         t.datetime :finished_at
         t.text :error
         t.integer :error_event, limit: 2
       end
   
       create_table :jobs_processes, id: :uuid do |t|
         t.timestamps
         t.jsonb :state
       end
   
       create_table :jobs_settings, id: :uuid do |t|
         t.timestamps
         t.text :key
         t.jsonb :value
         t.index :key, unique: true
       end
   
       add_index :jobs, :scheduled_at, where: "(finished_at IS NULL)", name: "index_jobs_on_scheduled_at"
       add_index :jobs, [:queue_name, :scheduled_at], where: "(finished_at IS NULL)", name: :index_jobs_on_queue_name_and_scheduled_at
       add_index :jobs, [:active_job_id, :created_at], name: :index_jobs_on_active_job_id_and_created_at
       add_index :jobs, :concurrency_key, where: "(finished_at IS NULL)", name: :index_jobs_on_concurrency_key_when_unfinished
       add_index :jobs, [:cron_key, :created_at], name: :index_jobs_on_cron_key_and_created_at
       add_index :jobs, [:cron_key, :cron_at], name: :index_jobs_on_cron_key_and_cron_at, unique: true
       add_index :jobs, [:active_job_id], name: :index_jobs_on_active_job_id
       add_index :jobs, [:finished_at], where: "retried_good_job_id IS NULL AND finished_at IS NOT NULL", name: :index_jobs_on_finished_at
       add_index :jobs, [:priority, :created_at], order: { priority: "DESC NULLS LAST", created_at: :asc },
         where: "finished_at IS NULL", name: :index_jobs_on_priority_created_at_when_unfinished
       add_index :jobs, [:batch_id], where: "batch_id IS NOT NULL"
       add_index :jobs, [:batch_callback_id], where: "batch_callback_id IS NOT NULL"
   
       add_index :jobs_executions, [:active_job_id, :created_at], name: :index_jobs_executions_on_active_job_id_and_created_at
     end
   end
   

4. Double-check your migration to make sure everything is updated.
5. Run the migration.
6. Your database should now contain the following tables:

   jobs
   jobs_batches
   jobs_executions
   jobs_processes
   jobs_settings
   

7. Restart the app.
8. And that should do it!