One-to-Many Patterns: Bucket, Array, Sub-collection - 008

Okka 'One-to-Many' problem ki 4 solutions unnay. Meeru eppudu 'Array of ObjectIds' eh vaduthunnara? Aithe mee app slow avvadam khayam.

Core Explanation:

1. Embedding: Few items (Addresses). Fast reads.

2. Referencing (Parent -> Child): Many items, but unbounded (Comments). Child holds Parent ID.

3. Array of References (Child -> Parent): Limited cardinality (Tags).

4. Bucket Pattern: Time-series or grouped data. Group many small docs into one medium doc.

Wrong Practice: Predicting array size wrong.

// Product
{
  "_id": "p1",
  "reviews": [oid1, oid2, ... oid10000?] // Dangerous
}

Best Practice: Child holds parent reference for unlimited growth.

// Review Document
{
  "_id": "r1",
  "product_id": "p1", // Reference to parent
  "text": "Good product"
}
db.reviews.createIndex({ product_id: 1 }) // Fast lookup

Closing Insight: "Few? Embed. Many? Child Reference. Huge? Bucket Pattern."