Mongodb

Description

MongoDB stores data records as BSON documents gathered in collections (= SQL table). BSON is a binary representation of JSON documents.
MongoDB documents are composed of field-and-value pairs. The value of a field can be any of the BSON data types.

var mydoc = { // field _id is a primary key _id: ObjectId("5099803df3f4948bd2f98391"), // name holds an embedded document that contains the fields first and last name: { first: "Alan", last: "Turing" }, birth: new Date('Jun 23, 1912'), death: new Date('Jun 07, 1954'), contribs: [ "Turing machine", "Turing test", "Turingery" ], views : NumberLong(1250000) }

Connect and authenticate

# connect mongosh mongodb://localhost:27017 # alias mongosh # connect and authenticate, use an environment variable to hide the password mongosh "mongodb://user:${DBPASSWORD}@<host>:<port>/admin?authSource=admin"

Authenticate

db.auth("username", "pwd") db.auth("username", passwordPrompt())

Database

// list the databases show dbs // switch to [dbname] use [dbname] // create a new database with a new collection (and insert data) use [newdbname] db.[newCollectionName].insertOne( { x: 1 } ) // drop the current database db.dropDatabase()

Collection

It gathers documents and is an equivalent of an SQL table.
A collection does not require its documents to have the same schema (the same set of fields and the same data types)

// list collections db.getCollectionInfos() db.runCommand('listCollections') db.createCollection("[collectionName]", { [option]: [value] })¨ // empty collection db.myCollection.deleteMany({})

View

It is a read-only queryable object whose contents are defined by an aggregation pipeline on other collections or views.
It is computed when you read the view, and it is not stored to disk.
Create and Query a View

// 2 syntaxes to create a view db.createCollection( "<viewName>", { "viewOn" : "<source>", "pipeline" : [<pipeline>], "collation" : { <collation> } } ) db.createView( "<viewName>", "<source>", [<pipeline>], { "collation" : { <collation> } } ) // drop a view db.<viewName>.drop()

Time Series Collections

MongoDB treats time series collections as writable non-materialized views backed by an internal collection. Create and Query a Time Series Collection

db.createCollection( "weather", { timeseries: { timeField: "timestamp", metaField: "metadata", granularity: "seconds" }})

Densify and fill

• densify allows to create missing spots in the TS
  • bounds: "full" spans from the full range of the original TS
  • bounds: [ <from>, <to> ] spans from the defined bounds
• fill allows to set values in those new spots.
  • locf to get the same value as the previous spot
  • linear to get the linear interpolation between the previous and the next point

[{ $densify: { field: "event", range: { step: 15, unit: "minute", bounds: "full" } } }, { $fill: { sortBy: { event: 1 }, output: { capacity: { method: "locf" } } } }]

Query

find

db.myCollection.find( { } ) // SELECT * db.myCollection.find( { item: 1, status: 1 } ) // SELECT _id, item, status // _id is returned by default db.myCollection.find( { item: 1, status: 1, _id: 0 } ) // SELECT item, status db.myCollection.find( { status: 0 } ) // returns all fields but status

filter (Query and Projection Operators)

db.myCollection.find( { name: "item1" } ) // SELECT ... WHERE name = "item1" db.myCollection.find( { name: /^item/ } ) // SELECT ... WHERE name LIKE "item%" db.myCollection.find( { status: "active", quantity: { $lt: 10 } } ) // SELECT ... WHERE status = "active" AND quantity < 10 db.myCollection.find( { $or: [ { status: "active" }, { qty: { $lt: 10 } } ] } ) // SELECT ... WHERE status = "active" OR quantity < 10 db.myCollection.find( { status: { $in: [ "active", "deleted" ] } } ) // SELECT ... WHERE status in ("active", "deleted")

• regex

Aggregation

db.myCollection.aggregate( [ ] )

match

$match : { name : "item1" }

sort

$match : { name : "item1" }

project

$project: { _id: 0, // hide _id property2: "$_id.property1", // include _id.property and rename as property2 property3: 1 // include property3 }

group

{ $group : { _id : "$item", totalAmount: { $sum: { $multiply: [ "$price", "$quantity" ] } } } }, { $match: { "totalAmount": { $gte: 100 } } } // GROUP BY item HAVING totalAmount > 100 $group: { _id: null, count: { $count: { } } } // SELECT COUNT(*) { $group: { _id: { state: "$state", city: "$city" }, cityTotalPopulation: { $sum: "$population" } } }, { $group: { _id: "$_id.state", avgCityPopulation: { $avg: "$cityTotalPopulation" } } } // SELECT SUM(population) AS cityTotalPopulation ... GROUP BY state, city // SELECT AVG(cityTotalPopulation) ... GROUP BY state // group and project { $group: { _id: { metadata: "$metadata", }, value: { $avg: "$value", }, count: { $count: {}, } } }, { $project: { _id: 0, metadata: "$_id.metadata", value: 1, count: 1 } }

densify

Creates new documents in a sequence of documents where certain values in a field are missing.

{ $densify: { field: "timestamp", partitionByFields: ["metadata", "value"], range: { step: 60, unit: "minute", bounds: [ new Date("2024-03-25T12:00:00.000Z"), new Date("2024-03-26T12:00:00.000Z"), ], }, }, }

fill

{ $fill: { sortBy: { timestamp: 1 }, partitionBy: { "metadata": "$metadata" }, output: { metadata: { method: "locf", }, value: { method: "locf", }, }, }, }

Index

// list indexes db.myCollection.getIndexes() // create an index on name db.myCollection.createIndex( { name: 1 } ) // for a single-field index, the sort order (ascending or descending) of the index key does not matter because MongoDB can traverse the index in either direction.

Backup and restore

# restore the backup folder dump to the local mongodb instance mongorestore dump/

don't know what to do with file while restore

This may happen because the mongorestore command point to a backup sub-folder not to the backup folder itself.

Authentication

// list the users of the current database db.getUsers() db.runCommand('usersInfo') // get info of a specific user db.getUser("tom", { showCredentials: true, showPrivileges: true, showAuthenticationRestrictions: true }) // create admin use admin db.createUser( { user: "admin", pwd: "mypwd", roles: [ { role: "userAdminAnyDatabase", db: "admin" } ] } ) // create a user for the demo database use demo db.createUser( { user: "tom", pwd: passwordPrompt(), // or cleartext password roles: [ { role: "readWrite", db: "demo" }, { role: "read", db: "finances" } ] } ) // change password use admin db.changeUserPassword("tom", "secretpassword") db.changeUserPassword("tom", passwordPrompt()) // delete a user db.dropUser("tom")

/etc/mongod.conf

security: authorization: "enabled"

Bash

mongo <<EOF use admin db.auth("admin", "${MONGODBADMINPWD}") use mydb db.dropDatabase() EOF

Installation

Windows

choco install mongodb mongodb-shell

Key	Value
Service name	MongoDB
Data directory	C:\Program Files\MongoDB\Server\6.0\data\
Log directory	C:\Program Files\MongoDB\Server\6.0\log\

• Install the Database Tools in C:\Program Files\MongoDB\Tools

Archlinux

yay mongodb-bin # install dependencies mongodb-bin-debug, mongosh-bin, mongosh-bin-debug # GUI yay mongodb-compass # tools yay mongodb-tools # service sc-status mongodb.service