Performance Challenges

Introduction

Performance challenges build the optimization reflex: spot the slow pattern, prove it with explain(), fix it with the smallest possible change, and measure the win.

Understanding the topic

6 timed performance fixes:

• 1. Convert a skip(10000) pagination to cursor-based.
• 2. Replace 100-field projection-less find with projected version.
• 3. Build a compound index that removes a SORT stage.
• 4. Rewrite an aggregation so $match happens before $lookup.
• 5. Move analytics reads to a tagged secondary.
• 6. Add a TTL index to bound an ever-growing sessions collection.

Informative example

Challenge 1 — skip → cursor:

// SLOW — skip becomes O(n) past page 100
db.posts.find({ status:"published" }).sort({ _id:-1 }).skip(10000).limit(20);

// FAST — cursor on _id, constant time
db.posts.find({ status:"published", _id: { $lt: lastSeenId } })
        .sort({ _id:-1 }).limit(20);

Best practices

• Measure before and after every change; without numbers it's superstition.
• Make one change at a time so you can attribute the win.
• Document each fix as a one-paragraph note — your future self will thank you.