spark 从数据库中读取数据

spark 从数据库中读取数据的几种方式，以MysqL 为例

创建spark

val spark = SparkSession.builer().appName("用户标签").master("local[*]").getorCreate()
val url = "jdbc:MysqL://数据库地址/数据库名"
val driver = "MysqL 数据库驱动”
val tableName = "表名"
val user = "数据库用户名"
val password = "数据库密码"

方式一: 直连

val df: DataFrame = spark.read.format("jdbc").options(Map(
      "url" -> url,
      "driver" -> driver
      "dbtable" -> tableName,
      "user"-> user,
      "password" -> password
      )).load()

方式二: 直连，执行指定sql

val sql = "select id, username from tableName"
val df: DataFrame = spark.read.format("jdbc").options(Map(
      "url" -> url,
      "driver" -> driver
      "dbtable" -> sql,
      "user"-> user,
      "password" -> password
      )).load()

方式三：使用jdbc 的properties

val prop = new Properties()
prop.setProperty("user", user)
prop.setProperty("password", password)
val df: DataFrame = spark.read.jdbc(url, tableName, prop)

方式四：传统的数据库驱动，查询结果是ResultSet 类型

try{
      Class.forName(driver)
var conn: Connection = DriverManager.getConnection(url, user, password)
val statement = conn.createStatement()
val res = statement.executeQuery(select * from tableName)
}

方式五：hive 数据库，与spark 天然结合（需要配置hive 作为数据源）

spark.sql("select * from tableName)