Spark-IoTDB

版本

Spark和Java所需的版本如下：

安装

mvn clean scala:compile compile install

Maven依赖

    <dependency>
      <groupId>org.apache.iotdb</groupId>
      <artifactId>spark-iotdb-connector</artifactId>
      <version>0.12.5</version>
    </dependency>

Spark-shell用户指南

spark-shell --jars spark-iotdb-connector-0.12.5.jar,iotdb-jdbc-0.12.5-jar-with-dependencies.jar
import org.apache.iotdb.spark.db._
val df = spark.read.format("org.apache.iotdb.spark.db").option("url","jdbc:iotdb://127.0.0.1:6667/").option("sql","select * from root").load
df.printSchema()
df.show()

如果要对rdd进行分区，可以执行以下操作

spark-shell --jars spark-iotdb-connector-0.12.0.jar,iotdb-jdbc-0.12.0-jar-with-dependencies.jar
import org.apache.iotdb.spark.db._
val df = spark.read.format("org.apache.iotdb.spark.db").option("url","jdbc:iotdb://127.0.0.1:6667/").option("sql","select * from root").
                        option("lowerBound", [lower bound of time that you want query(include)]).option("upperBound", [upper bound of time that you want query(include)]).
                        option("numPartition", [the partition number you want]).load
df.printSchema()
df.show()

模式推断

以下TsFile结构为例：TsFile模式中有三个度量：状态，温度和硬件。 这三种测量的基本信息如下：

TsFile中的现有数据如下：

• d1:root.ln.wf01.wt01
• d2:root.ln.wf02.wt02

宽（默认）表形式如下：

你还可以使用窄表形式，如下所示：（您可以参阅第4部分，了解如何使用窄表形式）

在宽和窄表之间转换

• 从宽到窄

import org.apache.iotdb.spark.db._
val wide_df = spark.read.format("org.apache.iotdb.spark.db").option("url", "jdbc:iotdb://127.0.0.1:6667/").option("sql", "select * from root where time < 1100 and time > 1000").load
val narrow_df = Transformer.toNarrowForm(spark, wide_df)

• 从窄到宽

import org.apache.iotdb.spark.db._
val wide_df = Transformer.toWideForm(spark, narrow_df)

Java用户指南

import org.apache.spark.sql.Dataset;
import org.apache.spark.sql.Row;
import org.apache.spark.sql.SparkSession;
import org.apache.iotdb.spark.db.*
public class Example {
  public static void main(String[] args) {
    SparkSession spark = SparkSession
        .builder()
        .appName("Build a DataFrame from Scratch")
        .master("local[*]")
        .getOrCreate();
    Dataset<Row> df = spark.read().format("org.apache.iotdb.spark.db")
        .option("url","jdbc:iotdb://127.0.0.1:6667/")
        .option("sql","select * from root").load();
    df.printSchema();
    df.show();
    Dataset<Row> narrowTable = Transformer.toNarrowForm(spark, df)
    narrowTable.show()
  }
}

写数据到IoTDB

用户指南

// import narrow table
val df = spark.createDataFrame(List(
      (1L, "root.test.d0",1, 1L, 1.0F, 1.0D, true, "hello"),
      (2L, "root.test.d0", 2, 2L, 2.0F, 2.0D, false, "world")))
val dfWithColumn = df.withColumnRenamed("_1", "Time")
    .withColumnRenamed("_2", "device_name")
    .withColumnRenamed("_3", "s0")
    .withColumnRenamed("_4", "s1")
    .withColumnRenamed("_5", "s2")
    .withColumnRenamed("_6", "s3")
    .withColumnRenamed("_7", "s4")
    .withColumnRenamed("_8", "s5")
dfWithColumn
    .write
    .format("org.apache.iotdb.spark.db")
    .option("url", "jdbc:iotdb://127.0.0.1:6667/")
    .save
// import wide table
val df = spark.createDataFrame(List(
      (1L, 1, 1L, 1.0F, 1.0D, true, "hello"),
      (2L, 2, 2L, 2.0F, 2.0D, false, "world")))
val dfWithColumn = df.withColumnRenamed("_1", "Time")
    .withColumnRenamed("_2", "root.test.d0.s0")
    .withColumnRenamed("_3", "root.test.d0.s1")
    .withColumnRenamed("_4", "root.test.d0.s2")
    .withColumnRenamed("_5", "root.test.d0.s3")
    .withColumnRenamed("_6", "root.test.d0.s4")
    .withColumnRenamed("_7", "root.test.d0.s5")
dfWithColumn.write.format("org.apache.iotdb.spark.db")
    .option("url", "jdbc:iotdb://127.0.0.1:6667/")
    .option("numPartition", "10")
    .save

注意

1. 无论dataframe中存放的是窄表还是宽表，都可以直接将数据写到IoTDB中。
2. numPartition参数是用来设置分区数，会在写入数据之前给dataframe进行重分区。每一个分区都会开启一个session进行数据的写入，来提高并发数。