Before 1970, there were flat file systems, problem : no standard implementation i.e. everybody had their own implementations so retrieval and insertion was a complication
In 1969, Relational databases provided a standard implementation But these could not handle big data ,they are not horizontally scalable (the ability to increase capacity by connecting multiple hardware or software entities so that they work as a single logical unit, keep on adding more and more computers to add more power) Performance is linearly proportional to number of computers
Recently, NoSQL Databases came into picture. Key-Value stores, Tablular database, Document oriented.
Horizontal scalability
sudo mkdir -p /data/db/
sudo chown id -u /data/db
export PATH=$PATH:{path upto bin}
type mongod in a terminal, this will give the message : “waiting for connections on port 27017”
Open a new terminal to connect to the server. Type mongo. This will start mongo shell.
show dbs
- shows all existing databases
use mycustomers
- command to create a new databse called my customers and switch to that database
db
- To check the current database
Sample JSON object : { first_name : “John”, last_name : “Doe”, memberships : [“mem1”,”mem2”], address : { street : “4 main st”, city : “Boston” } contacts : [ {name: “Brad”, relationship:”friend”} ] }
db.createUser({ user:”brad”, pwd:”1234”, roles:[“readWrite”,”dbAdmin”] });
Collections are very similar to tables in a database.
db.createCollection(‘customers’); show collections db.customers.insert({first_name:”John”, last_name:”Doe”});
to see documents in a collection :
db.customers.find(); ID field is automatically created
db.customers.insert([{first_name:”Steven”, last_name:”Smith”},{first_name:”Joan”, last_name:”Johnson”,gender:”female”}]); Even though we didn’t specify gender in the first 2 rows, the last one would have a field gender. i.e. we can add more fields as per our requirement
db.customers.find().pretty(); Displays the records in a better way
db.customers.update({first_name:”John”},{first_name:”John”,last_name:”Doe”,gender:”Male”}); Now John Doe will have a gender of male, if we missed a field it will be removed
db.customers.update({first_name:”Steven”},{$set:{gender:”Male”}) This will add a field. Previous fields will remain untouched.
inc operator :
db.customers.update({first_name:”Steven”},{$set:{age:”45”})
db.customers.update({first_name:”Steven”},{$inc:{age:5}});
operator unset :
db.customers.update({first_name:”Steven”,{($unset:{age:1}) Now steven will not have age field
db.customers.update({first_name:”Mary”,{first_name:”Mary”,last_name:”Samson”},{upsert : true}) with upsert the new updated row gets inserted
db.customer.remove({first_name:”Steven”}) Removes all the customers with first_name Steven
db.customer.remove({first_name:”Steven”},{justOne:true}) removes only one record
db.customers.find({first_name:”Sharon”}) => will give all sharon info
db.customers.find({$or:[first_name:”Sharon”, {first_name:”Troy”}]);
db.customers.find({age:{$lt:40}}).pretty(); gives every customer with age < 40
db.customers.find({“address.city”:Boston”}); There are some which would require quotes in the key as well
db.customers.find({memberships:”mem1”});
Sorting :
db.customers.find().sort({last_name:1}); Ascending order based on last_name for descending, set it to -1
db.customers.find().count(); db.customers.find({gender:”male”}).count(); db.customers.find().limit(4).sort({last_name:1});
Using for-each to iterate :
db.customers.find().forEach(function(doc){print(“Customer Name: “+doc.first_name)});