Let’s Make Data Management Easy ( SQL – No-SQL – CosmosDB)

Greetings to my Extra-Ordinary students across the GlobE .

Today I am very excited to talk about Database and specially Azure CosmosDB . Azure CosmosDB its a ultimate database solution from Microsoft and it’s belongs from NO-SQL family . So , before talk about Azure CosmosDB let me tell you something about Data , Data management and how it’s connected with SQL or No-SQL .

First of all I am not from DBA team and I had very less interaction with Databases . So at the beginning of this journey I faced couple of challenges to understand data model and differences between different data-store . So if you are not from DBA team please read and understand all the line very carefully . 

So I will start from very basic of Data and Data management and then I will talk about the advance part of it . In real world we work with huge no of data and that’s why we need a special software/application or infra to manage it .

SQL server is a product of Microsoft and through SQL server you can manage your data in structured way . To read and write data from SQL DB we have to follow its language or Structured Query Language .

Apart from Microsoft SQL server there are couple of Open-source options such as MySQL, PostgreSQL and SQLite . 

SQL means Structured Query Language and SQL DB(DataBase) is a place where keep our data and we can retrieve when we need it . SQL engine will do all task for us and we will get our expected result through query .

But what does DATA mean :- Name , Any Number like Phone Number – Aadhar Number , Address , Photo , Video anything that we want to save in our database and that is consider as a RAW data . But when we fetch it we get the data in Row and Column format that’s why this data consider is a Structured data or Process data.

Limitation : To store data in SQL at first we have to define the structure or schema or in easy term we have to create the TABLE . We can create this table through GUI or through CLI but we need Structured Query Language to fetch or read data .

Due to this limitation NO-SQL came in picture . So let’s talk about NO-SQL from scratch .

No-SQL is a schema free database . Means in NO-SQL we don’t need to define schema or the structure of table where data will store in background . Where in SQL at first we define the structure of table or define schema and then we feed the data . But in NO-SQL we do not define schema . 

So what is the benefit ? Benefit is less latency . Because its follow a simple instruction to find out the data or specially it doesn’t need and any special language to read or to fetch data .

To handle or to manage verity of data we need diff kind of database or data store that is provided by NO-SQL . Those category’s are :

1. Key-value
2. Document
3. Column-family 
4. Graph 

Based on your data type you have  to choose data-category . If you need all category then you may need to depend on multiple product like MongoDB, CouchDB, Redis and Apache Cassandra .

NoSQL databases have existed since the 1960s, but have been recently gaining traction with popular options such as MongoDB, CouchDB, Redis and Apache Cassandra .

I believe now you can understand the difference between SQL and NO-SQL . 

So let’s talk about CosmoDB .

CosmoDB is the only product that support multiple  data category . Means it support Key-Value , Document . And graph . It’s a multi API database . it means to feed the data you can use multiple solutions like MongoDB . 

Features of CosmoDB : 

• NoSQL JSON data Store
• Pushbutton geo-replication
• Low latency reads + writes
• Provisioned throughput
• Managed Service
• Multiple consistency choices
• No schemas , auto-indexed by default
• Multiple data models and languages

4 vector for SLA :

• availability
• latency
• consistency
• throughput

APIs for the following data models are supported with SDKs available in multiple languages:

• DocumentDB API: A schema-less JSON database engine with SQL querying capabilities.
• MongoDB API : A MongoDB database service built on top of Cosmos DB. Compatible with existing MongoDB libraries, drivers, tools and applications.
• Table API : A key-value storage database service built to provide premium capabilities for Azure Table storage applications.
• Graph (Gremlin) API : A graph database service built following the Apache TinkerPop specification.
• Additional data models coming soon! ** This information taken from Microsoft Official Website .

Let’s have a small comparison about different databases :

I believe now you understood the power of CosmoDB . So lets have a demo .

This is my Azure Subscription and here I am trying to create a Azure CosmoDB . So at first open portal and search for Azure CosmoDB .

Provide all the deatils and select one API for this database . It mean you can Migrate your any kind of database over here .

Here I am createing this database with MongoDB API .

As you can see my databse is provisioned and ready for work .

And this is the overview of CosmosDB .

Here is you all database connection string . You can copy this value and start working with your database .

If you want to use MongoDB API to migrate your data to Azure CosmosDB then you can

• Download either mongoimport.exe or mongorestore.exe from the MongoDB Download Center.
• 
• After provider all details you will be able to download this software .
• If you are importing data from MongoDB and plan to use it with the Azure Cosmos DB, you should use the Data Migration tool to import data.
• 
• 
• 

Apart from that you can use Robo3T (Formally known as RoboMongo) to migrate your data .

Now go ahed and download and install this software .

Now go back to Azure portal and open your CosmosDB Connection Strings tab . Copy the connection string URL and paste over there .

After providing all details click on TEST CONNECTION .

After this you will be connected with CosmosDB . 

It was basic understanding of No-SQL and Azure CosmosDB . This not the END of data and data management . We have different way to store and to mange the data .I will talk about Azure DataWareHouse and Hadoop or HDInsight very soon .

I will post something for DB Migration Very Soon . Till the time take care and bye bye .

** I would like to big thanks Kuldeep Kumar and Sujit Roy for the help .