N+1 QUERY ANALYSIS + FIX

PROBLEMATIC CODE (model: Post, has_many :comments)
   posts = Post.all
   posts.each do |post|
     post.comments.map(&:author)
   end
   Result: N+1: loads author for each comment

PROFILE: 1 initial query (posts) + 50 queries (1 per post for comments) = 51 queries

FIX 1: EAGER LOAD (simplest)
   posts = Post.includes(:comments).all
   posts.each do |post|
     post.comments.map(&:author)
   end
   
   Problem: Still N+1 for authors (loads all comment authors separately)
   New total: 2 queries (posts + comments loaded, but still separate author fetch)

FIX 2: DEEP EAGER LOAD (correct solution)
   posts = Post.includes(comments: :author).all
   
   Total queries: 3 (posts + comments + authors in one shot)
   Verify: Add .explain to confirm query plan

FIX 3: SELECT SPECIFIC COLUMNS (if you don't need all data)
   posts = Post.includes(:comments).select(:id, :title, :created_at)
   
   Benefit: Smaller payload, faster serialization

ACTIVE RECORD QUERY HINTS
   includes() → uses LEFT OUTER JOIN (fetches all, can be slow on huge joins)
   eager_load() → explicit JOIN (faster, used when filtering on association)
   preload() → separate queries (best when you don't need to filter on association)

PERFORMANCE MEASUREMENT
   Before: 51 queries, 800ms
   After (eager_load): 3 queries, 60ms
   Improvement: 13x faster

ADD INDEXES
   After queries are lean, add database index on Post.comments and Comment.author_id

RSPEC TEST (prevent regression)
   expect query count to not exceed 3 for the full operation