驱动下载

  创建maven工程,让maven来维护我们的jar,maven最重要的pom文件内容如下:

<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
    <modelVersion>4.0.0</modelVersion>

    <groupId>com.huawei</groupId>
    <artifactId>cassandra</artifactId>
    <version>0.0.1-SNAPSHOT</version>
    <packaging>jar</packaging>

    <name>cassandra</name>
    <url>http://maven.apache.org</url>

    <properties>
        <project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
    </properties>

    <dependencies>
       <dependency>
            <groupId>com.datastax.cassandra</groupId>
            <artifactId>cassandra-driver-core</artifactId>
            <version>2.1.10.3</version>
        </dependency>
    </dependencies>
</project>

Session获取

  官方给的Quick start案例

Cluster cluster = null;
        try {
            cluster = Cluster.builder()                                                    // (1)
                    .addContactPoint("127.0.0.1")                                 // cassandra服务器ip
                    .withCredentials("admin", "admin")                                     // 若没有启用账号认证,此处可以去掉
                    .build();
            Session session = cluster.connect();                                           // (2)

            ResultSet rs = session.execute("select release_version from system.local");    // (3)
            Row row = rs.one();
            System.out.println(row.getString("release_version"));                          // (4)
        } finally {
            if (cluster != null) cluster.close();                                          // (5)
        }

  代码中的(1) ~ (5)分别表示或者代表什么

    (1):Cluster对象是驱动程序的主入口点,它保存着真实Cassandra集群的状态(尤其是元数据);Cluster是线程安全的,一个Cassandra集群创建一个Cluster的单例,整个应用用这一个单例即可

    (2):Session用来执行查询的,而且它也是线程安全的,同样也应该重复利用

    (3):利用execute来发送一个查询到Cassandra,execute返回一个Resultset(结果集),这个结果集就是必要的列的行集合(二维表,行是满足条件的记录,列是我们关注的某些字段)

    (4):从row中提取数据

    (5):当任务完成后,关闭cluster,关闭cluster的同时将会关闭它创建的全部session;这一步很重要,它会释放潜在的资源(TCP连接、线程池等),在真实的应用中,我们应该在应用关闭(或应用卸载)的时候关闭cluster

    有jdbc开发的经验,就会发现,上述代码似曾相识,上述代码中的session就相当于jdbc中的connection,是整个数据库操作的基础,那么我们将session的获取单独抽出来

 
package com.huawei.cassandra.factory;

import java.util.concurrent.locks.Lock;
import java.util.concurrent.locks.ReentrantLock;

import com.datastax.driver.core.Cluster;
import com.datastax.driver.core.Session;

public class SessionRepository
{
    private static Session instance = null;
    private static Cluster cluster = null;
    private static Lock lock = new ReentrantLock();
    
    private SessionRepository(){}
    
    public static Session getSession()
    {
        if (null == instance)
        {
            try
            {
                lock.lock();
                
                if (null == instance)
                {
                    cluster = Cluster.builder()       
                            .addContactPoint("127.0.0.1")                
                            .withCredentials("admin", "admin")              
                            .build();
                    instance = cluster.connect();
                    // 也可以针对一个特定的keyspace获取一个session
                    // instance = cluster.connect("mycas");
                }
            }
            finally
            {
                lock.unlock();
            }
        }
        return instance;
    }
    
    public static void close()
    {
        if (null == cluster)
        {
            try
            {
                lock.lock();
                
                if (null == cluster)
                {
                    cluster.close();
                }
            }
            finally
            {
                lock.unlock();
            }
        }
    }
}

    拿到session了,那么请随意操作Cassandra吧!

cassandra基本操作

  创建表

    在mycas下创建表student

use mycas;
create table student(
    id int,
    address text,
    name text,
    age int,
    height int,
    primary key(id,address,name)
);

insert into student(id,address,name,age,height) values(1,'guangdong','lixiao',32,172);

  session直接执行cql

    和jdbc类似,关键是cql的拼接,下例是插入一条记录,删、改、查和这类似,不一一列举了

    // 字符串注意单引号'
        String cql = "insert into mycas.student(id,address,name,age,height) values(" 
                + student.getId() + ",'" + student.getAddress() + "','" + student.getName()
                + "'," + student.getAge() + "," + student.getHeight() + ");";
        System.out.println(cql);
        session.execute(cql);

  Querybuilder

    利用Querybuilder可以减轻cql的拼接,sql语句的拼接由驱动完成

    查询一个student:

    @Override
    public Student getStudentByKeys(int id, String address, String name)
    {
        Student student = null;
        ResultSet rs = session.execute(
                QueryBuilder.select("id", "address", "name", "age", "height")
                .from("mycas", "student")
                .where(QueryBuilder.eq("id", id))
                .and(QueryBuilder.eq("address", address))
                .and(QueryBuilder.eq("name", name)));
        Iterator<Row> rsIterator = rs.iterator();
        if (rsIterator.hasNext())
        {
            Row row = rsIterator.next();
            student = new Student();
            student.setAddress(row.getString("address"));
            student.setAge(row.getInt("age"));
            student.setHeight(row.getInt("height"));
            student.setId(row.getInt("id"));
            student.setName(row.getString("name"));
        }
        return student;
    }

    保存一个student:

  @Override
    public void saveStudent(Student student)
    {
        session.execute(
                QueryBuilder.insertInto("mycas", "student")
                .values(new String[]{"id", "address", "name", "age", "height"},
                        new Object[]{student.getId(), student.getAddress(), 
                            student.getName(), student.getAge(), student.getHeight()}));
    }

    修改一个student:

@Override
    public void updateStudent(Student student)
    {
        session.execute(
                QueryBuilder.update("mycas", "student")
                .with(QueryBuilder.set("age", student.getAge()))
                .and(QueryBuilder.set("height", student.getHeight()))
                .where(QueryBuilder.eq("id", student.getId()))
                .and(QueryBuilder.eq("address", student.getAddress()))
                .and(QueryBuilder.eq("name", student.getName())));
    }

    删除一个student:

@Override
    public void removeStudent(int id, String address, String name)
    {
        session.execute(QueryBuilder.delete()
                .from("mycas", "student")
                .where(QueryBuilder.eq("id", id))
                .and(QueryBuilder.eq("address", address))
                .and(QueryBuilder.eq("name", name)));
    }

  注意:驱动版本不同,Querybuilder的用法有些许不同,有些版本的某些方法变成非静态的了!

  占位符

    cassandra也有类似jdbc那样使用预编译占位符

    http://docs.datastax.com/en/developer/java-driver/3.0/manual/statements/prepared/

    预编译的原理是怎样的了,上面的链接是驱动官方的解释,我来谈谈我的理解

    当我们预编译statement的时候,Cassandra会解析query语句,缓存解析的结果并返回一个唯一的标志(PreparedStatement对象保持着这个标志的内部引用,就相当于通过标志可以获取到query语句预编译后的内容):

Cassandra-Java(增删查改)_apache

     当你绑定并且执行预编译statement的时候,驱动只会发送这个标志,那么Cassandra就会跳过解析query语句的过程:

Cassandra-Java(增删查改)_预编译_02

    所以,我们应该保证query语句只应该被预编译一次,缓存PreparedStatement 到我们的应用中(PreparedStatement 是线程安全的);如果我们对同一个query语句预编译了多次,那么驱动会打印警告日志;如果一个query语句只执行一次,那么预编译不会提供性能上的提高,反而会降低性能,因为它是两个来回(结合上面两张图),那么此时可以考虑用 simple statement 来代替

    和jdbc的预编译非常类似,我们来看看实际代码

    静态cql

  private static final String GET_STUDENT = "select id,address,name,age,height from mycas.student where id=? and address=? and name=?;";
  private static final String SAVE_STUDENT = "insert into mycas.student(id,address,name,age,height) values(?,?,?,?,?);";
  private static final String UPDATE_STUDENT = "update mycas.student set age=?, height=? where id=? and address=? and name=?;";
  private static final String REMOVE_STUDENT = "delete from mycas.student where id=? and address=? and name=?;";

    查询一个student

     Student student = null;
        PreparedStatement prepareStatement = session.prepare(GET_STUDENT);
        BoundStatement bindStatement = new BoundStatement(prepareStatement).bind(id, address, name);
        ResultSet rs = session.execute(bindStatement);
        Iterator<Row> rsIterator = rs.iterator();
        if (rsIterator.hasNext())
        {
            Row row = rsIterator.next();
            student = new Student();
            student.setAddress(row.getString("address"));
            student.setAge(row.getInt("age"));
            student.setHeight(row.getInt("height"));
            student.setId(row.getInt("id"));
            student.setName(row.getString("name"));
        }
        return student;

    保存一个student

        PreparedStatement prepareStatement = session.prepare(SAVE_STUDENT);
        BoundStatement bindStatement = new BoundStatement(prepareStatement)
            .bind(student.getId(), student.getAddress(), student.getName(), student.getAge(), student.getHeight());
        session.execute(bindStatement);

    修改一个student

     PreparedStatement prepareStatement = session.prepare(UPDATE_STUDENT);
        BoundStatement bindStatement = new BoundStatement(prepareStatement)
            .bind(student.getAge(), student.getHeight(), student.getId(), student.getAddress(), student.getName());
        session.execute(bindStatement);

    删除一个student

     PreparedStatement prepareStatement = session.prepare(REMOVE_STUDENT);
        BoundStatement bindStatement = new BoundStatement(prepareStatement)
            .bind(id, address, name);
        session.execute(bindStatement);

  批量batch

public static void batch()
    {
        Session session = SessionRepository.getSession();
        BoundStatement insertBind1 = new BoundStatement(
                session.prepare("insert into mycas.student(id,address,name,age,height) values(?,?,?,?,?);"))
                       .bind(3, "guangxi", "huangfeihong", 67, 175);
        
        BoundStatement insertBind2 = new BoundStatement(
                session.prepare("insert into mycas.student(id,address,name,age,height) values(?,?,?,?,?);"))
                       .bind(4, "hunan", "youzhibing", 26, 160);
        
        BoundStatement updateBind = new BoundStatement(
                session.prepare("update mycas.student set age=?, height=? where id=? and address=? and name=?;"))
                       .bind(72, 173, 3, "guangxi", "huangfeihong");
        
        BoundStatement deleteBind = new BoundStatement(
                session.prepare("delete from mycas.student where id=? and address=? and name=?;"))
                       .bind(4, "hunan", "youzhibing");
        
        BatchStatement batchStatement = new BatchStatement();
        batchStatement.add(insertBind1);
        batchStatement.add(insertBind2);
        batchStatement.add(updateBind);
        batchStatement.add(deleteBind);
        session.execute(batchStatement);
    }

public static void batch()
    {
        Session session = SessionRepository.getSession();
        BoundStatement insertBind1 = new BoundStatement(
                session.prepare("insert into mycas.student(id,address,name,age,height) values(?,?,?,?,?);"))
                       .bind(3, "guangxi", "huangfeihong", 67, 175);
        
        BoundStatement insertBind2 = new BoundStatement(
                session.prepare("insert into mycas.student(id,address,name,age,height) values(?,?,?,?,?);"))
                       .bind(4, "hunan", "youzhibing", 26, 160);
        
        BoundStatement updateBind = new BoundStatement(
                session.prepare("update mycas.student set age=?, height=? where id=? and address=? and name=?;"))
                       .bind(72, 173, 3, "guangxi", "huangfeihong");
        
        BoundStatement deleteBind = new BoundStatement(
                session.prepare("delete from mycas.student where id=? and address=? and name=?;"))
                       .bind(4, "hunan", "youzhibing");
        
        BatchStatement batchStatement = new BatchStatement();
        batchStatement.add(insertBind1);
        batchStatement.add(insertBind2);
        batchStatement.add(updateBind);
        batchStatement.add(deleteBind);
        session.execute(batchStatement);
    }