[R] R connect Hive with JDBC

最新推荐文章于 2023-08-12 05:53:52 发布

金乌儿

最新推荐文章于 2023-08-12 05:53:52 发布

阅读量209

点赞数 1

分类专栏： hands on exp

本文链接：https://blog.csdn.net/goyyue/article/details/84876120

版权

hands on exp 专栏收录该内容

4 篇文章 0 订阅

订阅专栏

#1. load packages

library("DBI")
library("rJava")
library("RJDBC")

#2. specify paths etc.

hive.class.path = list.files(path=c("/usr/local/hadoop/hive-release/lib"),pattern="jar", full.names=T);
hadoop.common.lib.path = list.files(path=c("/usr/local/hadoop/hadoop-2.7.1/share/hadoop/common"),pattern="jar",full.names=T);
hadoop.tool.lib.path = list.files(path=c("/usr/local/hadoop/hadoop-2.7.1/share/hadoop/tools/lib/"),pattern="jar",full.names=T);

cp = c(hive.class.path,hadoop.common.lib.path,hadoop.tool.lib.path)
.jinit(classpath=cp,parameters="-Djavax.security.auth.useSubjectCredsOnly=false")

#3. load the driver

#driver = JDBC(driver_class, driver_path)
drv <- JDBC("org.apache.hive.jdbc.HiveDriver", "/usr/local/hadoop/hive-release/lib/hive-jdbc-1.2.1-standalone.jar")

#3. define the connection

#connection = dbConnect(driver, address, user, password)
conn <- dbConnect(drv, "jdbc:hive2://xxx.xxx.x.x:port/xuserx;principal=hive/_HOST@xdomain;")

#4. query data

# if need to specify MapReduce queue
dbSendUpdate(conn, "set mapreduce.job.queuename=xxxx.xxxx.xxxx")
dbGetQuery(conn,"show tables")