Using #join with a Sidekiq::Batch in Rails to parallelize data migration

gistfile1.rb

class MoveInventoryUpdatesToS3 < ActiveRecord::Migration
  def up
    batch = Sidekiq::Batch.new
    batch.jobs do
      InventoryUpdate.where("content is not null").pluck(:id).each do |iuid|
        InventoryUpdate.delay.send_to_s3(iuid)
      end
    end
    
    # wait for all jobs to finish
    batch.status.join
  end
end

Beautiful, but what about a scenario where the migration adds a column (say for counter_cache: true) and we want to populate that using Sidekiq? I don't see anywhere to insert a InventoryUpdate.reset_column_information in your example.

That might be a case where you have a two-step migration or populate the column with a SQL UPDATE statement.