db.posts.find({ "likes" : {$gt : 100, $lte : 200}});

db.posts.find({ "likes" : { $gt : 100, $lt : 200}});

db.posts.find({ "likes" : { $gte : 100, $lt : 200}});

db.posts.find({ "likes" : { $gte : 100, $lte : 200}});

A btree index can be applied on any key of a document

A 2DSphere index can be applied on polygons data

2 indexes can be combined in a single query

A btree index behaves like a search engine (text queries)

{"rating" : 5}

{"val" : 4}

{"_id" : 1, "val" : 4}

{"_id" : ["Mark", "Hamill"], "val" : 4}

EncryptedStorage

WiredTiger

MMAPv1

InMemory

MongoFS

{ "_id" : 1, "firstname": ["George", "Junior"], "lastname" : "Lucas"}

{ "_id" : 1, "identity": {"firstname": "George", "lastname" : "Lucas"}}

{ "_id" : 1, "firstname": "George", "lastname" : "Lucas"}

{ "_id" : 1, "firstname": "george", "lastname" : {"name" : "Lucas"}}

{  "type" : "object",  

"additionalProperties" : false,  

"properties" : {  

"key1" : {"type": "string" },  

"key2" : {"type": "string" }  },  

"required" : ["key1"] }

{ "type" : "object", 

"properties" : { 

"key1" : {"type": "object" }, 

"key2" : {"type": "array" } }, 

"required" : ["key1"] }

{  "additionalProperties" : false,  

"properties" : { 

 "key1" : {"type": "string" },  

"key2" : {"type": "string" }  } }

{  "type" : "object",  

"properties" : {  

"key1" : {"type": "string" }, 

 "key2" : {"type": "number" }  }, 

 "required" : ["key1", "key2"] }

 "properties" : {  

"key1" : {"type": "string" },  

"key2" : {"type": "string" }  },  

"required" : ["key3"] }

All constraints remain consistent while applying a transaction at any moment

What ever happens the database gives an answer as fast as possible for each query

Any modification in the database cannot be reversed

Data will ultimately be up-to-date