🔍Lazy Loading in Mongoid: Understanding Mongoid::Criteria

When working with Mongoid (MongoDB ORM for Rails), you might write something like:

likes = Like.where(user: user)

At first glance, you might think this line runs a database query immediately. But, surprise, it doesn’t. In this post, we’ll unpack what’s happening behind the scenes with Mongoid::Criteria, how lazy loading works, and why it’s a performance-friendly feature you should love.

🧠 What is Mongoid::Criteria?

When you call:

likes = Like.where(user: user)

You're not retrieving any data yet. Instead, you get back an Mongoid::Criteria object, essentially a blueprint for the query.

puts likes.class
=> Mongoid::Criteria

This object describes what to fetch from the database, but Mongoid delays the actual DB call until it’s truly needed. This is what we mean by lazy loading.

🚀 When Does the Query Run?

The query only hits the database when something in your code forces evaluation. Common triggers include:

• .to_a

• .first

• .each

• .map

• .count

• .pluck

likes = Like.where(user: user) # Still just a query object
likes.each do |like|
puts like.likeable_type
end
=> Now the DB query actually runs here

⚙️ Why Lazy Loading Is Powerful

Lazy loading might feel a little magical at first, but it offers some serious benefits:

✅ Performance-friendly: Avoids unnecessary DB hits while you build your query
✅ Flexible: Chain additional conditions or scopes before executing
✅ Predictable (once you know it): Debugging gets easier when you understand the query lifecycle

🧪 How to See the Actual DB Query (Optional Debugging)

If you want to see exactly when the query runs, you can monitor your logs.

Rails.logger = Logger.new($stdout)
Rails.logger.level = 0
Mongo::Logger.logger= Rails.logger

Or you can just monitor your logs.

tail -f log/development.log

🧵 Wrap-up

Lazy loading isn’t unique to Mongoid—it’s a common pattern in many ORMs (ActiveRecord does it too). But understanding when your query runs is crucial for writing performant and bug-free code.