Read more

Use find_in_batches or find_each to deal with many records efficiently

Arne Hartherz
May 02, 2011Software engineer at makandra GmbH

Occasionally you need to do something directly on the server -- like having all records recalculate something that cannot be done in a migration because it takes a long time.

Illustration web development

Do you need DevOps-experts?

Your development team has a full backlog? No time for infrastructure architecture? Our DevOps team is ready to support you!

  • We build reliable cloud solutions with Infrastructure as code
  • We are experts in security, Linux and databases
  • We support your dev team to perform
Read more Show snapshot

Let's say you do something like this:


Even though you may have been successful with this on your development machine or the staging server, keep in mind that production machines often hold a lot more records. Using all may just work, even with lots of records, but when you iterate over such records and fetch associations for every object, those will be attached to them and kept in memory as well -- which causes more and more memory to be consumed over time.

When you need to process many records prefer find_in_batches Show snapshot :

Project.find_in_batches do |projects|
  projects.each do |project|

If the default of 1000 records per batch is still too much just set your own size like find_in_batches(:batch_size => 50).

When you only want to iterate in batches (and don't need to do anything on the scope find_in_batches gives you), you can use find_each Show snapshot (it does the each from above for you) as a shortcut:

Project.find_each do |project|

Both find_in_batches and find_each return records ordered by ID because they need to be able to iterate in batches. Modern Rails will raise an error if you try order yourself.

If you are on Rails 2.3, be aware that find calls inside the block are implicitly scoped Show snapshot . This is fixed in Rails 3+.

Posted by Arne Hartherz to makandra dev (2011-05-02 15:57)