Partitioning Technique in Parallel Database - Cassandra

This tutorial will help you to implement Partitioning Technique in Parallel Database like Cassandra.

Requirements:

1. Eclipse IDE
2. Cassandra install on Ubuntu.

Usefull Links:

• Installation of Eclipse on Ubuntu
• Installation of NoSQL Database - Cassandra

Steps:

1. Start the Cassandra Daemon using below code
~/cassandra/bin/cassandra -f
The cassandra daemon should start in the foreground (don’t press ctrl + c; as it’ll terminate the daemon)
2. Open eclipse -> new java project -> project name "Partitioning"
3. Create new class under your project -> and name the class "Partitioning"
4. Add the below code:

import java.util.Scanner;
import java.util.Vector;
import com.datastax.driver.core.Cluster;
import com.datastax.driver.core.ResultSet;
import com.datastax.driver.core.Row;
import com.datastax.driver.core.Session;

public class Partitioning {
    static Cluster cluster;
static Session session;
static ResultSet results;
static Row rows;
public static void main(String[] args)
{
//start the cassandra server in one terminal
try
{
cluster = Cluster.builder().addContactPoint("127.0.0.1").build();
session = cluster.connect();
session.execute("CREATE KEYSPACE IF NOT EXISTS keyspace1 WITH replication " + "= {'class':'SimpleStrategy','replication_factor':1}; ");
session.execute("USE keyspace1");
//create users table
    session.execute("CREATE TABLE IF NOT EXISTS users (lastname text, age int, city text, email text PRIMARY KEY, firstname text);");
    // Insert records into the users table
session.execute("INSERT INTO users (lastname, age, city, email, firstname) VALUES ('Jones', 35, 'Austin', 'bob@example.com', 'Bob')");
session.execute("INSERT INTO users (lastname, age, city, email, firstname) VALUES ('Swift', 25, 'Paris', 'mark@example.com', 'Mark')");
session.execute("INSERT INTO users (lastname, age, city, email, firstname) VALUES ('Castle', 35, 'New-York', 'kate@example.com', 'Kate')");
session.execute("INSERT INTO users (lastname, age, city, email, firstname) VALUES ('Jordon', 45, 'Mumbai', 'mike@example.com', 'Mike')");
session.execute("INSERT INTO users (lastname, age, city, email, firstname) VALUES ('Hanks', 48, 'Seoul', 'tom@example.com', 'Tom')");
session.execute("INSERT INTO users (lastname, age, city, email, firstname) VALUES ('Depp', 55, 'HongKong', 'johnny@example.com', 'Johnny')");
session.execute("INSERT INTO users (lastname, age, city, email, firstname) VALUES ('Nicholson', 65, 'Rio de Janeiro', 'jack@example.com', 'Jack')");
session.execute("INSERT INTO users (lastname, age, city, email, firstname) VALUES ('Freeman', 60, 'St Petersburg', 'morgan@example.com', 'Morgan')");
session.execute("INSERT INTO users (lastname, age, city, email, firstname) VALUES ('Lewis', 50, 'Alexandria', 'daniel@example.com', 'Daniel')");
session.execute("INSERT INTO users (lastname, age, city, email, firstname) VALUES ('Brando', 85, 'Ahmedabad', 'marlon@example.com', 'Marlon')");
session.execute("INSERT INTO users (lastname, age, city, email, firstname) VALUES ('DeNiro', 80, 'Toronto', 'robert@example.com', 'Robert')");
session.execute("INSERT INTO users (lastname, age, city, email, firstname) VALUES ('Bale', 50, 'Sydney', 'chris@example.com', 'Christian')");
//do partitioning 
    results = session.execute("SELECT * FROM users ");
int i=0,disk_no=0, total_disks=4;
System.out.println("\nRound Robin Partitioning");
    for (Row row : results) 
{ 
     /* Implement Round-Robin partitioning
Let, 'total_disks' = 4
disk_no = row_number % total_disks
*/
      disk_no=i++%total_disks;
  System.out.format("Disk: %d \t %s %d %s %s %s \n", 
disk_no,row.getString("lastname"), 
row.getInt("age"),row.getString("city"),row.getString("email"),row.getString("firstname"));
}
    int age;
    results = session.execute("SELECT * FROM users ");
    System.out.println("\nHash Partitioning");
    for (Row row : results) 
    {
  /* Implement Hash partitioning
Let, 'total_disks' = n = 4
suppose 'age' is the partitioning attribute
disk_no = (Partitioning attribute value) % total_disks
*/
  age=row.getInt("age");
  disk_no=age%total_disks;
  System.out.format("Disk: %d \t %s %d %s %s %s \n", 
disk_no,row.getString("lastname"), 
row.getInt("age"),row.getString("city"),row.getString("email"),row.getString("firstname"));
}
    int v,n;
    results = session.execute("SELECT * FROM users ");
 /* Implement Range Partitioning
suppose 'age' is the partitioning attribute
Let, 'total_disks' = n = 4
Let, partitioning vector : [vo, v1, ..., vn-2] = [v0, v1, v2] = [30,50,70]
Let, v be the partitioning attribute value of a tuple
Let, 0 <= i < n-2
Rule:
Tuples such that vi <= v < vi+1 go to disk i + 1. Tuples with v < v0 go to disk 0 and tuples
with v >= vn-2 go to disk n-1.
*/
      n= total_disks;
  int[] p_vector= new int[n-1];
  int flag;
  Scanner sc=new Scanner(System.in);
  System.out.println("\nRange Partitioning");
  System.out.println("\nPlease enter Partitioning vector values:-");
  for(i=0;i<=n-2;i++)
  {  
    System.out.println("Enter v"+i+" : ");
p_vector[i]=Integer.parseInt(sc.next());
  }
System.out.println("\nRange Partitioning - output \n");  
    for (Row row : results) 
    {
  v=row.getInt("age");
  flag=0;
  for(i=0;i<n-2;i++)
  {
  if(p_vector[i]<= v && v < p_vector[i+1])
  {
  disk_no=i+1;
  flag=1;
  break;
  }
  }
  if(flag==0)
  {
  if( v < p_vector[0])
  {
  disk_no=0;
  }  
  else if (v>= p_vector[n-2])
  {
  disk_no=n-1;
  }
  }
 System.out.format("Disk: %d \t %s %d %s %s %s \n", 
disk_no,row.getString("lastname"), 
row.getInt("age"),row.getString("city"),row.getString("email"),row.getString("firstname"));
}
sc.close();
}
catch(Exception e)
{
   System.out.println("Error: "+e.getMessage());
}
   cluster.close();
}
}

5.After adding the code it show errors to import Java Cassandra Driver
6. Download and Extract the Cassandra Java Driver - http://downloads.datastax.com/java-driver/cassandra-java-driver-2.0.2.tar.gz
7. Add the jar file into your projects
Go to Eclipse -> Package Explorer(Left Box)-> Partitioning(project name) -> JRE System Library -> Right click -> Build Path -> Configure Build Path
8. Click on Add External Jars -> Go to cassandra-java-driver-2.0.2 folder -> Add the jar files
Note: There are total 12 jar files. 2 jar files in cassandra-java-driver-2.0.2 and 10 jar files in cassandra-java-driver-2.0.2 -> lib.
9. Done, Now run the project