【Spark-SQL学习之二】 SparkSQL DataFrame创建和储存
环境
虚拟机:VMware 10
Linux版本:CentOS-6.5-x86_64
客户端:Xshell4
FTP:Xftp4
jdk1.8
scala-2.10.4(依赖jdk1.8)
spark-1.6
1、读取json格式的文件创建DataFrame
注意:
(1)json文件中的json数据不能嵌套json格式数据。
(2)DataFrame是一个一个Row类型的RDD,df.rdd()/df.javaRdd()。
(3)可以两种方式读取json格式的文件。
sqlContext.read().format(“json”).load(“path”)
sqlContext.read().json(“path”)
(4)df.show()默认显示前20行数据。
(5)DataFrame原生API可以操作DataFrame(不方便)。
(6)注册成临时表时,表中的列默认按ascii顺序显示列。
数据:json
{"name":"zhangsan","age":"20"}
{"name":"lisi"}
{"name":"wangwu","age":"18"}
示例代码:
Java:
package com.wjy.df; import org.apache.spark.SparkConf; import org.apache.spark.SparkContext; import org.apache.spark.api.java.JavaRDD; import org.apache.spark.rdd.RDD; import org.apache.spark.sql.DataFrame; import org.apache.spark.sql.Row; import org.apache.spark.sql.SQLContext; /** * 读取json格式的文件创建DataFrame * * 注意 :json文件中不能嵌套json格式的内容 * * 1.读取json格式两种方式 * 2.df.show默认显示前20行,使用df.show(行数)显示多行 * 3.df.javaRDD/(scala df.rdd) 将DataFrame转换成RDD * 4.df.printSchema()显示DataFrame中的Schema信息 * 5.dataFram自带的API 操作DataFrame ,用的少 * 6.想使用sql查询,首先要将DataFrame注册成临时表:df.registerTempTable("jtable"),再使用sql,怎么使用sql?sqlContext.sql("sql语句") * 7.不能读取嵌套的json文件 * 8.df加载过来之后将列按照ascii排序了 * @author root * */ public class CreateDFFromJosonFile { public static void main(String[] args) { SparkConf conf = new SparkConf().setMaster("local").setAppName("CreateDFFromJosonFile"); SparkContext sc = new SparkContext(conf);//注意 这里不是JavaSparkContext //创建SQLContext SQLContext sqlContext = new SQLContext(sc); /** * DataFrame的底层是一个一个的RDD RDD的泛型是Row类型。 * 以下两种方式都可以读取json格式的文件 * {"name":"zhangsan","age":"20"} {"name":"lisi"} {"name":"wangwu","age":"18"} */ DataFrame df = sqlContext.read().format("json").load("./data/json");//{"name":"zhangsan","age":"20"}; df.show();// 显示 DataFrame中的内容,默认显示前20行。如果现实多行要指定多少行show(行数) 注意:当有多个列时,显示的列先后顺序是按列的ascii码先后显示。 DataFrame df2 = sqlContext.read().json("./data/json"); df2.show(); /* * +----+--------+ | age| name| +----+--------+ | 20|zhangsan| |null| lisi| | 18| wangwu| +----+--------+ */ //DataFrame转换成RDD JavaRDD<Row> javaRDD = df.javaRDD(); //树形的形式显示schema信息 df.printSchema(); /* * root |-- age: string (nullable = true) |-- name: string (nullable = true) */ //dataFram自带的API 操作DataFrame 这种方式比较麻烦 用的比较少 //select name from table df.select("name").show(); /* * +--------+ | name| +--------+ |zhangsan| | lisi| | wangwu| +--------+ */ //select name ,age+10 as addage from table df.select(df.col("name"),df.col("age").plus(10).alias("addage")).show(); /* * +--------+------+ | name|addage| +--------+------+ |zhangsan| 30.0| | lisi| null| | wangwu| 28.0| +--------+------+ */ //select name ,age from table where age>19 df.select(df.col("name"),df.col("age")).where(df.col("age").gt(19)).show(); /* * +--------+---+ | name|age| +--------+---+ |zhangsan| 20| +--------+---+ */ //select age,count(*) from table group by age df.groupBy(df.col("age")).count().show(); /* * +----+-----+ | age|count| +----+-----+ | 18| 1| | 20| 1| |null| 1| +----+-----+ */ //将DataFrame注册成临时的一张表,这张表相当于临时注册到内存中,是逻辑上的表,不会物化到磁盘 这种方式用的比较多 df.registerTempTable("person"); DataFrame df3 = sqlContext.sql("select age,count(*) as gg from person group by age"); df3.show(); DataFrame df4 = sqlContext.sql("select age, name from person"); df4.show(); sc.stop(); } }
Scala:
package com.wjy.df import org.apache.spark.SparkConf import org.apache.spark.SparkContext import org.apache.spark.sql.SQLContext object CreateDFFromJsonFile { def main(args:Array[String]):Unit={ val conf = new SparkConf().setMaster("local").setAppName("CreateDFFromJsonFile"); val sc = new SparkContext(conf); val sqlContext = new SQLContext(sc); val df1 = sqlContext.read.json("./data/json"); df1.show(); val df2 = sqlContext.read.format("json").load("./data/json"); df2.show(); val rdd = df1.rdd; df1.printSchema(); //select name from table df1.select(df1.col("name")).show(); //select name from table where age>19 df1.select(df1.col("name"),df1.col("age")).where(df1.col("age").gt(19)).show(); //select count(*) from table group by age df1.groupBy(df1.col("age")).count().show(); //注册临时表 df1.registerTempTable("person"); val df3 = sqlContext.sql("select * from person"); df3.show(); /* * +----+--------+ | age| name| +----+--------+ | 20|zhangsan| |null| lisi| | 18| wangwu| +----+--------+ */ sc.stop(); } }
2、通过json格式的RDD创建DataFrame
RDD的元素类型是String,但是格式必须是JSON格式
示例代码:
Java:
package com.wjy.df; import java.util.Arrays; import org.apache.spark.SparkConf; import org.apache.spark.SparkContext; import org.apache.spark.api.java.JavaRDD; import org.apache.spark.api.java.JavaSparkContext; import org.apache.spark.sql.DataFrame; import org.apache.spark.sql.SQLContext; public class CreateDFFromJsonRDD { public static void main(String[] args) { SparkConf conf = new SparkConf().setMaster("local").setAppName("CreateDFFromJsonRDD"); //SparkContext sc = new SparkContext(conf); JavaSparkContext sc = new JavaSparkContext(conf); SQLContext sqlContext = new SQLContext(sc); JavaRDD<String> javaRDD1 = sc.parallelize(Arrays.asList("{'name':'zhangsan','age':\"18\"}", "{\"name\":\"lisi\",\"age\":\"19\"}", "{\"name\":\"wangwu\",\"age\":\"20\"}")); JavaRDD<String> javaRDD2 = sc.parallelize(Arrays.asList("{\"name\":\"zhangsan\",\"score\":\"100\"}", "{\"name\":\"lisi\",\"score\":\"200\"}", "{\"name\":\"wangwu\",\"score\":\"300\"}")); DataFrame namedf = sqlContext.read().json(javaRDD1); namedf.show(); DataFrame scoredf = sqlContext.read().json(javaRDD2); scoredf.show(); //DataFrame原生API使用 //SELECT t1.name,t1.age,t2.score from t1, t2 where t1.name = t2.name namedf.join(scoredf, namedf.col("name").$eq$eq$eq(scoredf.col("name"))) .select(namedf.col("name"),namedf.col("age"),scoredf.col("score")).show(); //注册成临时表 namedf.registerTempTable("name"); scoredf.registerTempTable("score"); //如果自己写的sql查询得到的DataFrame结果中的列会按照 查询的字段顺序返回 DataFrame result = sqlContext.sql("select name.name,name.age,score.score " + "from name join score " + "on name.name = score.name"); result.show(); /* * +--------+---+-----+ | name|age|score| +--------+---+-----+ |zhangsan| 18| 100| | lisi| 19| 200| | wangwu| 20| 300| +--------+---+-----+ */ sc.stop(); } }
Scala:
package com.wjy.df import org.apache.spark.SparkConf import org.apache.spark.SparkContext import org.apache.spark.sql.SQLContext object CreateDFFromJsonRDD { def main(args:Array[String]):Unit={ val conf = new SparkConf().setMaster("local").setAppName("CreateDFFromJsonRDD"); val sc = new SparkContext(conf); val sqlContext = new SQLContext(sc); val rdd1 = sc.makeRDD(Array( "{\"name\":\"zhangsan\",\"age\":18}", "{\"name\":\"lisi\",\"age\":19}", "{\"name\":\"wangwu\",\"age\":20}" )); val rdd2 = sc.makeRDD(Array( "{\"name\":\"zhangsan\",\"score\":100}", "{\"name\":\"lisi\",\"score\":200}", "{\"name\":\"wangwu\",\"score\":300}" )); val namedf = sqlContext.read.json(rdd1); val scoredf = sqlContext.read.json(rdd2); namedf.registerTempTable("name"); scoredf.registerTempTable("score"); val result = sqlContext.sql("select name.name,name.age,score.score from name,score where name.name = score.name"); result.show(); sc.stop(); } }
3、通过非json格式的RDD来创建出来一个DataFrame
(1)通过反射的方式 (不建议使用)
(1.1)自定义类要可序列化(注意变量被关键字transient修饰 则不会被序列化;静态变量也不能被序列化)
注意ava中以下几种情况下不被序列化的问题:
1.1.1.反序列化时serializable 版本号不一致时会导致不能反序列化。
1.1.2.子类中实现了serializable接口,父类中没有实现,父类中的变量不能被序列化,序列化后父类中的变量会得到null。
注意:父类实现serializable接口,子类没有实现serializable接口时,子类可以正常序列化
1.1.3.被关键字transient修饰的变量不能被序列化。
1.1.4.静态变量不能被序列化,属于类,不属于方法和对象,所以不能被序列化。
另外:一个文件多次writeObject时,如果有相同的对象已经写入文件,那么下次再写入时,只保存第二次写入的引用,读取时,都是第一次保存的对象。
(1.2)自定义类的访问级别是Public
(1.3)RDD转成DataFrame后会根据映射将字段按Assci码排序
(1.4)将DataFrame转换成RDD时获取字段两种方式,一种是df.getInt(0)下标获取(不推荐使用),另一种是df.getAs(“列名”)获取(推荐使用)
示例代码:
Java:
package com.wjy.df; import org.apache.spark.SparkConf; import org.apache.spark.api.java.JavaRDD; import org.apache.spark.api.java.JavaSparkContext; import org.apache.spark.api.java.function.Function; import org.apache.spark.api.java.function.VoidFunction; import org.apache.spark.sql.DataFrame; import org.apache.spark.sql.Row; import org.apache.spark.sql.SQLContext; /** * @author Administrator * 通过反射的方式将非json格式的RDD转换成DataFrame * 注意:这种方式不推荐使用 */ public class CreateDFFromRDDWithReflect { public static void main(String[] args) { SparkConf conf = new SparkConf().setMaster("local").setAppName("CreateDFFromRDDWithReflect"); JavaSparkContext sc = new JavaSparkContext(conf); SQLContext sqlContext = new SQLContext(sc); /* * 1,zhansan,18 2,lisi,19 3,wangwu,20 */ JavaRDD<String> lineRDD = sc.textFile("./data/person.txt"); JavaRDD<Person> personRDD = lineRDD.map(new Function<String, Person>() { private static final long serialVersionUID = 1L; @Override public Person call(String line) throws Exception { String[] ss = line.split(","); Person p = new Person(); p.setId(ss[0]); p.setName(ss[1]); p.setAge(Integer.valueOf(ss[2])); return p; } }); /** * 传入进去Person.class的时候,sqlContext是通过反射的方式创建DataFrame * 在底层通过反射的方式获得Person的所有field,结合RDD本身,就生成了DataFrame */ DataFrame df1 = sqlContext.createDataFrame(personRDD, Person.class); df1.show(); df1.printSchema(); df1.registerTempTable("person"); DataFrame ret = sqlContext.sql("select name,id,age from person where id = 2"); ret.show(); /* * +----+---+---+ |name| id|age| +----+---+---+ |lisi| 2| 19| +----+---+---+ */ /** * 将DataFrame转成JavaRDD * 注意: * 1.可以使用row.getInt(0),row.getString(1)...通过下标获取返回Row类型的数据,但是要注意列顺序问题---不常用 * 2.可以使用row.getAs("列名")来获取对应的列值。 */ JavaRDD<Row> javaRDD = ret.javaRDD(); JavaRDD<Person> map = javaRDD.map(new Function<Row, Person>() { private static final long serialVersionUID = 1L; @Override public Person call(Row row) throws Exception { //顺序和ret一致 Person p = new Person(); // p.setId(row.getString(1)); // p.setName(row.getString(0)); // p.setAge(row.getInt(2)); p.setId(row.getAs("id")); p.setName(row.getAs("name")); p.setAge(row.getAs("age")); return p; } }); map.foreach(new VoidFunction<Person>() { private static final long serialVersionUID = 1L; @Override public void call(Person p) throws Exception { System.out.println(p); } }); sc.stop(); } }
Scala:
package com.wjy.df import org.apache.spark.SparkConf import org.apache.spark.SparkContext import org.apache.spark.sql.SQLContext //case class 默认是可以序列化的,也就是实现了Serializable;ase class构造函数的参数是public级别 case class Person(id:String,name:String,age:Integer); object CreateDFFromRDDWithReflect { def main(args:Array[String]):Unit={ val conf = new SparkConf().setMaster("local").setAppName("CreateDFFromRDDWithReflect"); val sc = new SparkContext(conf); val sqlContext = new SQLContext(sc); val lineRDD = sc.textFile("./data/person.txt"); val personRDD = lineRDD.map { x => { val p = Person(x.split(",")(0), x.split(",")(1), Integer.valueOf(x.split(",")(2))); p }}; //将RDD隐式转换成DataFrame import sqlContext.implicits._ val df = personRDD.toDF(); df.show(); /* * +---+-------+---+ | id| name|age| +---+-------+---+ | 1|zhansan| 18| | 2| lisi| 19| | 3| wangwu| 20| +---+-------+---+ */ //DataFrame转成RDD val rdd = df.rdd; val result = rdd.map { x => { Person(x.getAs("id"),x.getAs("name"),x.getAs("age")); }}; result.foreach {println}; /* * Person(1,zhansan,18) Person(2,lisi,19) Person(3,wangwu,20) */ sc.stop(); } }
(2)动态创建schema的方式
示例代码:
Java:
package com.wjy.df; import java.util.Arrays; import java.util.List; import org.apache.spark.SparkConf; import org.apache.spark.api.java.JavaRDD; import org.apache.spark.api.java.JavaSparkContext; import org.apache.spark.api.java.function.Function; import org.apache.spark.sql.DataFrame; import org.apache.spark.sql.Row; import org.apache.spark.sql.RowFactory; import org.apache.spark.sql.SQLContext; import org.apache.spark.sql.types.DataTypes; import org.apache.spark.sql.types.StructField; import org.apache.spark.sql.types.StructType; /** * @author Administrator * * 动态创建Schema将非json格式RDD转换成DataFrame */ public class CreateDFFromRDDWithStruct { public static void main(String[] args) { SparkConf conf = new SparkConf().setMaster("local").setAppName("CreateDFFromRDDWithStruct"); JavaSparkContext sc = new JavaSparkContext(conf); SQLContext sqlContext = new SQLContext(sc); JavaRDD<String> lineRDD = sc.textFile("./data/person.txt"); //转换成Row类型的RDD JavaRDD<Row> rowrdd = lineRDD.map(new Function<String, Row>() { private static final long serialVersionUID = 1L; @Override public Row call(String line) throws Exception { String[] ss = line.split(","); return RowFactory.create(ss[0],ss[1],Integer.valueOf(ss[2])); } }); //动态构建DataFrame中的元数据,一般来说这里的字段可以来源自字符串,也可以来源于外部数据库 List<StructField> asList = Arrays.asList( DataTypes.createStructField("id", DataTypes.StringType, true), DataTypes.createStructField("name", DataTypes.StringType, true), DataTypes.createStructField("age", DataTypes.IntegerType, true) ); //根据元数据创建schema StructType schema = DataTypes.createStructType(asList); //根据row和schema创建DataFrame DataFrame df = sqlContext.createDataFrame(rowrdd, schema); df.show(); /* * +---+-------+---+ | id| name|age| +---+-------+---+ | 1|zhansan| 18| | 2| lisi| 19| | 3| wangwu| 20| +---+-------+---+ */ sc.stop(); } }
Scala:
package com.wjy.df import org.apache.spark.SparkConf import org.apache.spark.SparkContext import org.apache.spark.sql.SQLContext import org.apache.spark.sql.RowFactory import org.apache.spark.sql.types.StructType import org.apache.spark.sql.types.StringType import org.apache.spark.sql.types.StructField import org.apache.spark.sql.types.IntegerType object CreateDFFromRDDWithStruct { def main(args:Array[String]):Unit={ val conf = new SparkConf().setMaster("local").setAppName("CreateDFFromRDDWithStruct"); val sc = new SparkContext(conf); val sqlContext = new SQLContext(sc); val lineRDD = sc.textFile("./data/person.txt"); //row val rowRDD = lineRDD.map { x => { val ss = x.split(","); RowFactory.create(ss(0),ss(1),Integer.valueOf(ss(2))); }}; //schema val schema = StructType(List( StructField("id",StringType,true), StructField("name",StringType,true), StructField("age",IntegerType,true) )); //根据row和schema创建DataFrame val df = sqlContext.createDataFrame(rowRDD, schema); df.show(); sc.stop(); } }
4、读取parquet文件创建DF
注意:
可以将DataFrame存储成parquet文件。保存成parquet文件的方式有两种
df.write().mode(SaveMode.Overwrite)format("parquet").save("./sparksql/parquet");
df.write().mode(SaveMode.Overwrite).parquet("./sparksql/parquet");
SaveMode指定文件保存时的模式。
Overwrite:覆盖
Append:追加
ErrorIfExists:如果存在就报错
Ignore:如果存在就忽略
示例代码:
Java:
package com.wjy.df; import org.apache.spark.SparkConf; import org.apache.spark.api.java.JavaRDD; import org.apache.spark.api.java.JavaSparkContext; import org.apache.spark.sql.DataFrame; import org.apache.spark.sql.SQLContext; import org.apache.spark.sql.SaveMode; public class CreateDFFromParquet { public static void main(String[] args) { SparkConf conf = new SparkConf().setMaster("local").setAppName("CreateDFFromParquet"); JavaSparkContext sc = new JavaSparkContext(conf); SQLContext sqlContext = new SQLContext(sc); JavaRDD<String> jsonRDD = sc.textFile("./data/json"); DataFrame dataFrame = sqlContext.read().json(jsonRDD); dataFrame.show(); /** * 将DataFrame保存成parquet文件, * SaveMode指定存储文件时的保存模式: * Overwrite:覆盖 * Append:追加 * ErrorIfExists:如果存在就报错 * Ignore:如果存在就忽略 * 保存成parquet文件有以下两种方式: */ //方式一:save dataFrame.write().mode(SaveMode.Overwrite).format("parquet").save("./data/parquet"); //方式二:parquet dataFrame.write().mode(SaveMode.Ignore).parquet("./data/parquet"); /* * Initialized Parquet WriteSupport with Catalyst schema: { "type" : "struct", "fields" : [ { "name" : "age", "type" : "string", "nullable" : true, "metadata" : { } }, { "name" : "name", "type" : "string", "nullable" : true, "metadata" : { } } ] } and corresponding Parquet message type: message spark_schema { optional binary age (UTF8); optional binary name (UTF8); } */ /** * 加载parquet文件成DataFrame * 加载parquet文件有以下两种方式: */ //方式一:load DataFrame df1 = sqlContext.read().format("parquet").load("./data/parquet"); df1.show(); //方式二:parquet DataFrame df2 = sqlContext.read().parquet("./data/parquet"); df2.show(); sc.stop(); } }
Scala:
package com.wjy.df import org.apache.spark.SparkConf import org.apache.spark.SparkContext import org.apache.spark.sql.SQLContext import org.apache.spark.sql.SaveMode object CreateDFFromParquet { def main(args:Array[String]):Unit={ val conf = new SparkConf().setMaster("local").setAppName("CreateDFFromParquet"); val sc = new SparkContext(conf); val sqlContext = new SQLContext(sc); val jsonRDD = sc.textFile("./data/json"); val df = sqlContext.read.json(jsonRDD); df.show(); /** * 将DF保存为parquet文件 */ df.write.mode(SaveMode.Overwrite).format("parquet").save("./data/parquet"); df.write.mode(SaveMode.Ignore).parquet("./data/parquet"); /** * 读取parquet文件 */ val df1 = sqlContext.read.format("parquet").load("./data/parquet"); df1.show(); val df2 = sqlContext.read.parquet("./data/parquet"); df.show(); sc.stop(); } }
5、读取JDBC中的数据创建DataFrame(MySql为例)
两种方式创建DataFrame
第一种方式读取MySql数据库表,加载为DataFrame
第二种方式读取MySql数据表加载为DataFrame
示例代码:
Java:
package com.wjy.df; import java.util.HashMap; import java.util.Map; import java.util.Properties; import org.apache.spark.SparkConf; import org.apache.spark.api.java.JavaSparkContext; import org.apache.spark.sql.DataFrame; import org.apache.spark.sql.DataFrameReader; import org.apache.spark.sql.SQLContext; import org.apache.spark.sql.SaveMode; public class CreateDFFromMysql { public static void main(String[] args) { SparkConf conf = new SparkConf().setMaster("local").setAppName("CreateDFFromMysql"); /** * 配置join或者聚合操作shuffle数据时分区的数量 */ conf.set("spark.sql.shuffle.partitions", "1"); JavaSparkContext sc = new JavaSparkContext(conf); SQLContext sqlContext = new SQLContext(sc); /** * 第一种方式读取MySql数据库表,加载为DataFrame */ Map<String, String> options = new HashMap<String,String>(); options.put("url", "jdbc:mysql://134.32.123.101:3306/spark"); options.put("driver", "com.mysql.jdbc.Driver"); options.put("user", "root"); options.put("password", "123456"); options.put("dbtable", "person"); DataFrame df1 = sqlContext.read().format("jdbc").options(options).load(); df1.show(); df1.registerTempTable("person1"); /** * 第二种方式读取MySql数据表加载为DataFrame */ DataFrameReader reader = sqlContext.read().format("jdbc"); reader.option("url", "jdbc:mysql://134.32.123.101:3306/spark"); reader.option("driver", "com.mysql.jdbc.Driver"); reader.option("user", "root"); reader.option("password", "123456"); reader.option("dbtable", "score"); DataFrame df2 = reader.load(); df2.show(); df2.registerTempTable("score1"); DataFrame dataFrame = sqlContext.sql("select person1.id,person1.name,person1.age,score1.score " + "from person1,score1 " + "where person1.name = score1.name"); dataFrame.show(); /** * 将DataFrame结果保存到Mysql中 */ Properties properties = new Properties(); properties.setProperty("user", "root"); properties.setProperty("password", "123456"); /** * SaveMode: * Overwrite:覆盖 * Append:追加 * ErrorIfExists:如果存在就报错 * Ignore:如果存在就忽略 * */ dataFrame.write().mode(SaveMode.Overwrite).jdbc("jdbc:mysql://134.32.123.101:3306/spark", "result", properties); System.out.println("----Finish----"); sc.stop(); } }
Scala:
package com.wjy.df import org.apache.spark.SparkConf import org.apache.spark.SparkContext import org.apache.spark.sql.SQLContext import java.util.HashMap import java.util.Properties import org.apache.spark.sql.SaveMode object CreateDFFromMysql { def main(args:Array[String]):Unit={ val conf = new SparkConf().setMaster("local").setAppName("CreateDFFromMysql"); val sc = new SparkContext(conf); val sqlContext = new SQLContext(sc); /** * 第一种方式读取Mysql数据库表创建DF */ val options = new HashMap[String,String](); options.put("url", "jdbc:mysql://134.32.123.101:3306/spark") options.put("driver","com.mysql.jdbc.Driver") options.put("user","root") options.put("password", "123456") options.put("dbtable","person") val df1 = sqlContext.read.format("jdbc").options(options).load(); df1.show(); df1.registerTempTable("person"); /** * 第二种方式读取Mysql数据库表创建DF */ var reader = sqlContext.read.format("jdbc"); reader.option("url", "jdbc:mysql://134.32.123.101:3306/spark") reader.option("driver","com.mysql.jdbc.Driver") reader.option("user","root") reader.option("password","123456") reader.option("dbtable", "score") val df2 = reader.load(); df2.show(); df2.registerTempTable("score"); val result = sqlContext.sql("select person.id,person.name,score.score from person,score where person.name = score.name") result.show() /** * 将数据写入到Mysql表中 */ val properties = new Properties() properties.setProperty("user", "root") properties.setProperty("password", "123456") result.write.mode(SaveMode.Overwrite).jdbc("jdbc:mysql://134.32.123.101:3306/spark", "result", properties); sc.stop(); } }
6、读取Hive中的数据加载成DataFrame
HiveContext是SQLContext的子类,连接Hive建议使用HiveContext。
由于本地没有Hive环境,要提交到集群运行,提交命令:
./spark-submit --master spark://node1:7077,node2:7077 --executor-cores 1 --executor-memory 2G --total-executor-cores 1 --class com.bjsxt.sparksql.dataframe.CreateDFFromHive /root/test/HiveTest.jar
示例代码:
Java:
package com.wjy.df; import org.apache.spark.SparkConf; import org.apache.spark.api.java.JavaSparkContext; import org.apache.spark.sql.DataFrame; import org.apache.spark.sql.Row; import org.apache.spark.sql.SaveMode; import org.apache.spark.sql.hive.HiveContext; /** * 如果读取hive中数据,要使用HiveContext * HiveContext.sql(sql)可以操作hive表,还可以操作虚拟的表 * */ public class CreateDFFromHive { public static void main(String[] args) { //不能设置local了 需要打成jar在hive上运行 SparkConf conf = new SparkConf().setAppName("CreateDFFromHive"); JavaSparkContext sc = new JavaSparkContext(conf); //HiveContext是SQLContext的子类。 使用hive sql操作 HiveContext hiveContext = new HiveContext(sc); hiveContext.sql("USE Spark");//使用spark数据库 //表student_infos hiveContext.sql("drop table if exists student_infos");//删除表 hiveContext.sql("CREATE TABLE IF NOT EXISTS student_infos (name STRING,age INT) row format delimited fields terminated by '\t' ");//创建表 hiveContext.sql("load data local inpath '/root/test/student_infos' into table student_infos");//hive加载数据 //表student_scores hiveContext.sql("DROP TABLE IF EXISTS student_scores"); hiveContext.sql("CREATE TABLE IF NOT EXISTS student_scores (name STRING, score INT) row format delimited fields terminated by '\t'"); hiveContext.sql("LOAD DATA LOCAL INPATH '/root/test/student_scores' INTO TABLE student_scores"); /** * 查询表生成DataFrame */ DataFrame student_infos = hiveContext.table("student_infos"); student_infos.show(); DataFrame student_scores = hiveContext.table("student_scores"); student_scores.show(); DataFrame goodStudentsDF = hiveContext.sql("SELECT si.name, si.age, ss.score " + "FROM student_infos si " + "JOIN student_scores ss " + "ON si.name=ss.name " + "WHERE ss.score>=80"); goodStudentsDF.show(); goodStudentsDF.registerTempTable("goodStudent"); DataFrame result = hiveContext.sql("select * from goodstudent"); result.show(); /** * 将结果保存到hive表 good_student_infos */ hiveContext.sql("DROP TABLE IF EXISTS good_student_infos"); goodStudentsDF.write().mode(SaveMode.Overwrite).saveAsTable("good_student_infos"); DataFrame table = hiveContext.table("good_student_infos"); Row[] rows = table.collect(); for (Row row:rows) { System.out.println(row); } sc.stop(); } }
Scala:
package com.wjy.df import org.apache.spark.SparkConf import org.apache.spark.SparkContext import org.apache.spark.sql.hive.HiveContext import org.apache.spark.sql.SaveMode object CreateDFFromHive { def main(args:Array[String]):Unit={ //依赖hive 不能设置local模式 val conf = new SparkConf().setAppName("CreateDFFromHive"); val sc = new SparkContext(conf); /** * HiveContext是SQLContext的子类。 */ val hiveContext = new HiveContext(sc); hiveContext.sql("use spark") //student_infos hiveContext.sql("drop table if exists student_infos") hiveContext.sql("create table if not exists student_infos (name string,age int) row format delimited fields terminated by '\t'") hiveContext.sql("load data local inpath '/root/test/student_infos' into table student_infos") val df1 = hiveContext.table("student_infos"); df1.show(); //student_scores hiveContext.sql("drop table if exists student_scores") hiveContext.sql("create table if not exists student_scores (name string,score int) row format delimited fields terminated by '\t'") hiveContext.sql("load data local inpath '/root/test/student_scores' into table student_scores") val df2 = hiveContext.table("student_scores"); df2.show(); val df = hiveContext.sql("select si.name,si.age,ss.score from student_infos si,student_scores ss where si.name = ss.name") df.show(); /** * 将结果写入到hive表中 */ //good_student_infos hiveContext.sql("drop table if exists good_student_infos") df.write.mode(SaveMode.Overwrite).saveAsTable("good_student_infos"); sc.stop(); } }
附:Spark On Hive的配置
1、在Spark客户端配置Hive On Spark
在Spark客户端安装包下spark-1.6.0/conf中创建文件hive-site.xml:
配置hive的metastore路径
<configuration> <property> <name>hive.metastore.uris</name> <value>thrift://node1:9083</value> </property> </configuration>
2、启动Hive的metastore服务
hive --service metastore
3、启动zookeeper集群,启动HDFS集群,启动spark集群。
4、启动SparkShell 读取Hive中的表总数,对比hive中查询同一表查询总数测试时间。
./spark-shell --master spark://node1:7077,node2:7077 --executor-cores 1 --executor-memory 1g --total-executor-cores 1 ...... scala>import org.apache.spark.sql.hive.HiveContext; scala>val hc = new HiveContext(sc); scala>hc.sql("show databases").show(); scala>hc.sql("user default").show(); scala>hc.sql("select count(*) from jizhan").show();
注意:
如果使用Spark on Hive 查询数据时,出现错误:Caused by:java.net.UnkonwnHostException:....
找不到HDFS集群路径,要在客户端机器conf/spark-env.sh中设置HDFS的路径:
export HADOOP_CONF_DIR=$HADOOP_HOME/etc/hadoop
参考:
Spark