Who this is for

This chapter covers the core features and capabilities of MongoDB Operations Toolkit.

Who this is for

Backend engineers and DBAs who already use MongoDB and want correct, opinionated

patterns for the decisions that are easy to get wrong: embed vs. reference, the

ESR index rule, choosing a shard key you will not regret, reading an explain

plan, and taking a backup you can actually restore.

Prerequisites

• MongoDB 6.0+ (7.0 recommended). The aggregation stages used here

($setWindowFields, $densify, $merge) need 5.0+; $lookup on sharded

collections needs 5.1+. Where a feature needs a newer server, the doc says so.

• mongosh (the modern shell — the legacy mongo shell is not supported).
• Database Tools (mongodump, mongorestore) on your PATH for the backup

scripts. They ship separately from the server as mongodb-database-tools.

• A MongoDB you can connect to. Scripts default to

mongodb://127.0.0.1:27017; override with the MONGO_URI environment variable.

Hostnames such as mongo-rs0-01.example.com are placeholders — replace them

with your own. No real infrastructure, addresses, or credentials appear

anywhere in this product.

How to run

Follow this guide to get MongoDB Operations Toolkit up and running in your environment.

How to run

# 1. Load the sample data so every example has something to query
mongosh "mongodb://127.0.0.1:27017/shop" examples/sample_schema.js

# 2. Build the indexes the queries rely on
mongosh "mongodb://127.0.0.1:27017/shop" scripts/create_indexes.js

# 3. Run the aggregation pipelines (sales rollups, RFM, funnel, top-N)
mongosh "mongodb://127.0.0.1:27017/shop" scripts/aggregation_pipelines.js

# 4. Turn on the profiler and inspect slow operations
mongosh "mongodb://127.0.0.1:27017/shop" scripts/profiling_queries.js

# 5. Check cluster health
mongosh "mongodb://127.0.0.1:27017" scripts/replica_set_status.js
mongosh "mongodb://127.0.0.1:27017" scripts/shard_status.js

# 6. Take a backup
bash shell/mongodump_backup.sh

# 7. (Optional) spin up a local 3-node replica set to practice against
bash shell/replica_set_init.sh

mongosh file.js runs a script against a connection. Inside an interactive

shell you can instead load('scripts/aggregation_pipelines.js').

Contents

mongodb-operations-toolkit/
├── README.md                       This file
├── LICENSE                         MIT
├── config/
│   └── mongod.conf                 Annotated: WiredTiger cache, journal, profiler, repl, security
├── docs/
│   ├── schema-design-patterns.md   Embed vs reference + 6 named patterns (subset, bucket, outlier, ...)
│   ├── indexing-guide.md           Compound order (ESR), partial/TTL/text, covered queries, $regex traps
│   ├── replica-sets.md             Elections, read/write concern, oplog sizing, lag, maintenance
│   ├── sharding.md                 Shard-key selection, hashed vs ranged, chunks, balancer, anti-patterns
│   ├── performance-profiling.md    Profiler levels, explain() plans, the slow-query workflow
│   └── backup-and-restore.md       mongodump/restore, filesystem snapshots, PITR, restore drills
├── scripts/
│   ├── create_indexes.js           Index builders with rationale per index (ESR applied)
│   ├── aggregation_pipelines.js    Six worked pipelines on the sample data
│   ├── profiling_queries.js        Enable profiler, find slow ops, COLLSCANs, unused indexes
│   ├── replica_set_status.js       Health, lag, oplog window, election readiness report
│   └── shard_status.js             Shard distribution, chunk balance, jumbo-chunk check
├── shell/
│   ├── mongodump_backup.sh         Timestamped dump + gzip + retention + restore note
│   └── replica_set_init.sh         Bring up a local 3-node replica set for practice
└── examples/
    ├── sample_schema.js            Creates collections + validators and inserts sample docs
    └── sample_data.json            The raw sample documents (orders, users, products, events)