/LearnMongoDB

Learn MongoDB

What is data?

  • The facts that can be stored or recorded.

What is database?

  • The group of data that can be relate together.

OLAP & OLTP & NewSQL

  • OLTP : Online transactional processing : All those transaction takes place in the real-time/Dynamic data, provdies short term storage. Ex: OLTP => RDBMS => Oracle
  • OLAP : Offline analytical processing : Access all the processed/static or historical data, provdies long term storage. Ex: OLAP => DW => Netezza
  • NewSQL : Stores all application data, provdies long term storage. NoSQL => MongoDB

Evolution of Database

  1. Flat file (IBM- Programmable sequencial file) 1960's Store and retrive information any time. Can't filter the data &cannot have a table structure there.
  2. Hierarchey Database There is relation between each entity. As structure becomes bigger, it's difficult to get relation between higher and lower chain structure.
  3. DBMS 1970's
    • RDBMS : Entity based relationship model.
      • Volume : It's designed to work with concentrated data, ie., limited data.
      • Variety : It's designed to worked with structure/semi-strcutured data.
  4. NoSQL database
    • VVV - Volume,Variety, Velocity
    • Volume : Huge storage of data,It's designed to work with distributed data.
    • Variety : It's designed to worked with unstructure data.
    • Velocity : NoSQL provides velocity becuase of it's flexibility with volume and variety.
    • It was introduced to overcome drawback with SQL
    • It has no schema or table structure
    • It has distributed data.
    • It support Replication of data
    • It also support horizantal scaling of data ie., Sharding
    • It also open source.
    • NoSQL stores value in form of either
      1. key-value pairs,
      2. wide-column
      3. graph
      4. Documents.
    • MongoDB is used to store JSON Documents.
      • Google : Bog Table & Amazon : Amazon DynamoDB => Initial NoSQL database

How to decide which database to use?

ACID ? BASE ? CAP ?

Properties :

  • RDBMS works on ACID properties:
  1. ACID : Atomicity, Consistency,Isolation & Durability
    • Atomicity : Database will commit if all the transcaction are successful, if any of them fails, then roll-back the commit.
    • Consistency : Both parties involved in transaction must see same changes at the end of the transaction.
    • Isolation : When two transaction is taking place,each transaction should be individual in their own.
    • Durability : All the changes must be permanent, they should be volatile in nature.
  • NoSQL works on BASE properties:

  1. BASE: Basically Available, Soft State & Eventually consistent

    • Basically Available : Database is up & running, but won't be able to support all operations. Ex: when server is down, transaction might not be possibl;e but one can see their status.
    • Soft State : Able to move from one state to another state, It doesn't has to be write-consistent or all mutally consistent all the time.
    • Eventually consistent : Same changes appeares eventually over a peroid of time. Ex: When one upload 300 pic on facebook, eventually it appears to all friends.

  2. CAP Theorm : According to CAP theorm, no way any database in the world can display these 3 properties at same time, Consistency, Availabity, Partition Tolerance (Horizontal data scaling => 1TB + 1TB) Ex: MongoDb supports Consistency, Partition Tolerance RDBMS supports Availabity, Consistency Cassendra supports Availabity, Partition Tolerance

MongoDB Database

  • It supports Consistency, Partition Tolerance
  • It is also highly scablable
  • It's highly Available,but not completely Available(Basically availability)