MongoDB
The document database for flexible, evolving schemas.
MongoDB is a document-oriented NoSQL database that stores data as BSON (binary JSON). It's the default choice for applications with rapidly evolving schemas, deeply nested data, and high-volume writes — where rigid SQL DDL would slow development down.
Deploy with Pier
- 1 Open the Pier dashboard and click Add service.
- 2 Pick MongoDB from the template list.
- 3 Choose the version, set a service name, and Pier provisions the container, storage, and ports automatically.
- 4 Attach a domain if you want HTTPS. Traefik auto-provisions the Let's Encrypt certificate.
What is MongoDB?
MongoDB is the most-deployed NoSQL document database. Instead of tables and rows, it stores BSON documents — nested objects much like JSON — grouped into collections. Schemas are optional; queries use a rich JSON-like syntax that maps naturally to JavaScript and other dynamic languages.
The trade-off vs SQL is real: you give up cross-collection foreign-key integrity and SQL-standard ANSI compliance. In exchange you get schema-on-read flexibility, deep document atomicity, and a query language that fits the shape of modern app data better than rigid relational rows.
How Pier deploys it
Pier uses the official mongo Docker image, mounting /data/db as the
persistent volume. The default version is latest (MongoDB 8.x); 7 and 6
are available for strict version pinning. Root credentials are auto-generated
via MONGO_INITDB_ROOT_USERNAME + MONGO_INITDB_ROOT_PASSWORD.
Single-node only in this template — multi-node replica sets require manual
initialization and aren’t supported via Pier’s cluster mode yet. Backups use
mongodump --archive --gzip on schedule and can push to S3.
When NOT to use MongoDB
For workloads with rigid relational integrity needs (financial ledgers, inventory with strict foreign keys) — pick Postgres. For full-text search — Elasticsearch. For embedded single-process — SQLite. Mongo shines when your data has nested structure that changes shape often and you want JSON-native ergonomics.
Key features
Schema flexibility
Every document in a collection can have its own shape. Add fields without ALTER TABLE; remove them without migrations. Validation rules can be enforced when needed.
Powerful aggregation pipeline
$match, $group, $lookup (join), $facet, $bucket, $unwind, $graphLookup — MongoDB's aggregation framework rivals SQL for analytical queries on document data.
Native horizontal scaling
Sharding by hashed or ranged key splits collections across nodes. Replica sets give automatic failover, read scaling, and zero-downtime maintenance.
Geospatial indexes
2dsphere indexes for GeoJSON points, polygons, multipolygons. $near, $geoWithin, $geoIntersects queries with kilometer-level accuracy.
Change streams (CDC)
Subscribe to inserts/updates/deletes in real time via change streams. Build reactive UIs, sync to other systems, drive event-driven workflows.
Time series collections (5.0+)
Dedicated time series collections with columnar storage and automatic bucketing — purpose-built for IoT and metrics workloads.
Use cases
User-generated content platforms
Social media posts, comments, reactions — flexible per-document fields fit naturally into BSON.
Real-time IoT data ingest
High write throughput, time-series collections, and change streams make Mongo a viable purpose-built IoT backend.
Product catalog with variable attributes
E-commerce catalogs with different attributes per category (TVs have screen size; books have ISBN) — no need for sparse SQL columns or EAV.
Mobile app sync backend
Mongo Realm + Atlas Device Sync. For self-hosted, pair with WatermelonDB or PouchDB-style sync layers.
Single view / 360 customer profiles
Aggregate data from many systems into one denormalized document per customer. Atomic updates on nested fields.
Code examples
db.users.insertOne({
email: "[email protected]",
profile: {
name: "Aleksandr",
locale: "en",
tags: ["beta", "pro"]
},
createdAt: new Date()
});db.users.aggregate([
{ $unwind: "$profile.tags" },
{ $group: {
_id: { locale: "$profile.locale", tag: "$profile.tags" },
count: { $sum: 1 }
}},
{ $sort: { count: -1 } },
{ $limit: 20 }
]);db.places.createIndex({ location: "2dsphere" });
db.places.find({
location: {
$near: {
$geometry: { type: "Point", coordinates: [37.6173, 55.7558] },
$maxDistance: 5000
}
}
});const stream = db.orders.watch([
{ $match: { "fullDocument.status": "paid" } }
]);
stream.on("change", (ev) => sendToWarehouse(ev.fullDocument));How it compares
| vs PostgreSQL with JSONB | Postgres's JSONB matches Mongo's flexibility while keeping joins, transactions, and SQL. If you need both relational data and documents in one place, Postgres often wins. Pick Mongo when documents are the primary model and you want change streams + sharding native. |
| vs CouchDB | CouchDB is master-master with offline-first sync — great for mobile. Mongo wins on aggregation power, ecosystem, and tooling. |
| vs Elasticsearch | Elastic is purpose-built for full-text search; Mongo is for general document storage. They often pair — Mongo for source of truth, Elastic for search indices. |
| vs DynamoDB | Dynamo is AWS-managed and pay-per-request; Mongo is self-hosted with predictable costs. Mongo has aggregations Dynamo lacks. |
Frequently asked questions
Which MongoDB version does Pier deploy?
Is replication supported?
Does the SSPL license affect me?
How do I connect from my app?
Default authentication?
Are backups automatic?
When should I add indexes?
Related services
Deploy on your VPS
MongoDB is a document-oriented NoSQL database that stores data as BSON (binary JSON). It's the default choice for applications with rapidly evolving schemas, deeply nested data, and high-volume writes — where rigid SQL DDL would slow development down.
Deploy this service →