原文地址:http://www.ithao123.cn/content-2523824.html
[摘要:正在build.sbt中设置装备摆设依附(止之间须要空) ame := test2 scalaVersion := 2.10.4 libraryDependencies = Seq( org.apache.spark % spark-core % 1.0.0, org.apache.hbase % hbase % 1.2.1-hadoop1, org.apache.hbase % hbase-client % 1.2.1-hado]
在build.sbt中配置依赖(行之间需要空格)
ame := "test2"
scalaVersion := "2.10.4"
libraryDependencies ++= Seq(
"org.apache.spark" % "spark-core" % "1.0.0",
"org.apache.hbase" % "hbase" % "1.2.1-hadoop1",
"org.apache.hbase" % "hbase-client" % "1.2.1-hadoop1",
"org.apache.hbase" % "hbase-common" % "1.2.1-hadoop1",
"org.apache.hbase" % "hbase-server" % "1.2.1-hadoop1"
)
version := "1.0"
在环境变量中配置该工程需要的JAR包,命名为SPARK_TEST_JAR。
————————————————————————————————————————
import org.apache.hadoop.hbase.client._
import org.apache.hadoop.hbase.mapreduce.TableInputFormatimport org.apache.hadoop.hbase.util.Bytes
import org.apache.hadoop.hbase._
import org.apache.spark._
/**
* Created by gongxuan on 2/3/15.
* hadoop1.2.1 scala2.10.4 hbase0.98.9 spark1.0.0
*
*/
object HBaseTest {
def main(args: Array[String]) {
//create table test_table1
var table_name = "test1"
val conf = HBaseConfiguration.create
val admin = new HBaseAdmin(conf)
if (admin.tableExists(table_name))
{
admin.disableTable(table_name)
admin.deleteTable(table_name)
}
val htd = new HTableDescriptor(table_name)
val hcd = new HColumnDescriptor("id")
//add column to table
htd.addFamily(hcd)
admin.createTable(htd)
//put data to HBase table
val tablename = htd.getName
val table = new HTable(conf, tablename)
val databytes = Bytes.toBytes("id")
for (c <- 1 to 10) {
val row = Bytes.toBytes("row" + c.toString)
val p1 = new Put(row)
p1.add(databytes, Bytes.toBytes(c.toString), Bytes.toBytes("value" + c.toString))
table.put(p1)
}
for (c <- 1 to 10) {
val g = new Get(Bytes.toBytes("row" + c.toString))
println("Get:" + table.get(g))
}
//search table
val config = HBaseConfiguration.create
val sc = new SparkContext("local", "HBaseTest",
System.getenv("SPARK_HOME"), Seq(System.getenv("SPARK_TEST_JAR")))
config.set(TableInputFormat.INPUT_TABLE,table_name)
// 用hadoopAPI创建一个RDD
val hbaseRDD = sc.newAPIHadoopRDD(config, classOf[TableInputFormat],classOf[org.apache.hadoop.hbase.io.ImmutableBytesWritable],
classOf[org.apache.hadoop.hbase.client.Result])
val count = hbaseRDD.count()
println("HbaseRDD Count:" + count)
hbaseRDD.cache()
//找到result对象 ,返回类型Array[(org.apache.hadoop.hbase.io.ImmutableBytesWritable,org.apache.hadoop.hbase.client.Result)]
//key是一个不变的ImmutableBytesWritable,value是Hbase的Result
//res(0)._2返回第二个参数result
val res = hbaseRDD.take(count.toInt)for (j <- 1 to count.toInt) {
var rs = res(j - 1)._2
//遍历res.raw取出每一个单元的值 ,返回类型:Array[org.apache.hadoop.hbase.KeyValue]
var kvs = rs.rawfor (kv <- kvs) // 再遍历每一个单元里面的记录
println("row:" + new String(kv.getRow()) +
" cf:" + new String(kv.getFamily()) +
" column:" + new String(kv.getQualifier()) +
" value:" + new String(kv.getValue()))
}
//drop table
admin.disableTable(table_name)
admin.deleteTable(table_name)
}
}