🍃 MongoDB
The document store people mean when they say NoSQL — embed or reference, the 16MB limit and the unbounded array that takes a year to appear, the missing $set that replaces your whole document, aggregation pipelines, and a shard key you only choose once.
No lessons published yet — this is the roadmap, so you can see exactly what the course will cover before it lands.
Before you start: NoSQL Concepts — Series 1, and Series 2. Query-first modelling and the embed-or-reference decision are taught there, once, for every non-relational engine. This track applies them to one product and does not re-argue them.
Short on time? The 16 episodes marked core are the shortest path to being productive at work. Start there and pick up the rest whenever a job asks for them.
Series 1 · Documents and Queries
CRUD, the aggregation pipeline, and reading an explain
- 01
Documents, Collections and BSON — What Replaces the Row and the Table
coreintermediatecoming soon - 02
find() — the Filter Is a Document, and Arrays Match Differently
coreintermediatecoming soon - 03
Update Operators — and the Missing $set That Replaces the Whole Document
coreintermediatecoming soon - 04
The Aggregation Pipeline — Stages in Order, and Filtering Early
coreintermediatecoming soon - 05
Indexes and explain() — the ESR Rule, and docsExamined Against nReturned
coreintermediatecoming soon - 06
Project — Answer Five Questions From a Collection You Did Not Design
coreintermediate#projectcoming soon
Series 2 · Modelling for MongoDB
embed, reference, and the patterns that keep documents sane
- 01
Embed or Reference — One-to-Few, One-to-Many, One-to-Squillions
coreadvancedcoming soon - 02
16MB and the Unbounded Array — the Antipattern That Takes a Year to Appear
coreadvancedcoming soon - 03
The Patterns — Extended Reference, Computed, Bucket, Subset and Versioning
coreadvancedcoming soon - 04
Schema Validation — Schemaless Was Never a Property of Your Application
coreadvancedcoming soon - 05
$lookup Is Not a JOIN — and Several of Them Is a Modelling Message
coreadvancedcoming soon - 06
Project — Take a Relational Schema and Model It as MongoDB Wants It
coreadvanced#projectcoming soon
Series 3 · In Production
replica sets, shard keys, transactions and change streams
- 01
Replica Sets — Write Concern, Read Preference, and Reading Your Own Writes
coreadvancedcoming soon - 02
Sharding — the Shard Key Decides Everything, and Most Deployments Need None
coreadvancedcoming soon - 03
Multi-Document Transactions — They Exist, and They Are Not the First Answer
advancedcoming soon - 04
Change Streams — CDC Built In, and the Oplog Window That Limits It
coreadvancedcoming soon - 05
The Working Set — Why RAM Is the Number, and the Profiler That Proves It
advancedcoming soon - 06
Project — Review a Shard Key Against Its Real Queries and Say What Breaks
coreadvanced#projectcoming soon
Not here, on purpose
- Why a document store at all, and how it differs from the other shapes — covered in NoSQL Concepts, Series 1 and Series 2, which this course assumes
- Distributed systems theory — quorums, CAP, eventual consistency — covered in NoSQL Concepts, Series 3. In this course, Series 3 · 00 is what read and write concern do in this product specifically
- Running the server — deployment, upgrades, monitoring, capacity — covered in Data Ops for the discipline. That track runs PostgreSQL, so the habits transfer and the commands do not
- Storing JSON inside a relational database — covered in Learn SQL, Series 3 · 05 for querying it and Learn Data Modeling, Series 2 · 42 for deciding to. Often the better answer, and NoSQL Concepts, Series 4 · 00 argues that case
- Atlas as a managed platform, and Atlas-only features — not covered beyond naming them. Search, Vector Search and Charts are a hosted product's surface, and an episode about them would age faster than anything else here