optimal_batch_size.rb

def self.find_optimal_batch_size
  times = []

  [100, 1_000, 10_000, 100_000].each do |batch_size|
    times << [timed_query(batch_size), batch_size]
  end

  times.sort
end

def self.timed_query(batch_size)
  max_batches = (LIMIT / BATCH_SIZE.to_f).ceil

  Benchmark.realtime do
    batch_count = 0
    govcloud_company_ids = Company.where(use_govcloud: true).select(:id)

    ProstoreFile.unscoped.select(:id).where(company_id: govcloud_company_ids).
    find_in_batches(batch_size: batch_size) do |prostore_files|
      ProstoreFile.unscoped.where(id: prostore_files.map(&:id)).update_all(storage_profile_key: nil)
      batch_count +=1
      break if batch_count == max_batches
    end

    batch_count = 0

    ProstoreFile.unscoped.select(:id).where.not(company_id: govcloud_company_ids).
    find_in_batches(batch_size: batch_size) do |prostore_files|
      ProstoreFile.unscoped.where(id: prostore_files.map(&:id)).update_all(storage_profile_key: nil)
      batch_count +=1
      break if batch_count == max_batches
    end

    batch_count = 0

    ProstoreFile.unscoped.select(:id).where(company_id: nil).
    find_in_batches(batch_size: batch_size) do |prostore_files|
      ProstoreFile.unscoped.where(id: prostore_files.map(&:id)).update_all(storage_profile_key: nil)
      batch_count +=1
      break if batch_count == max_batches
    end
  end
end