2016-04-26 209 views
1

這是Save Spark dataframe as dynamic partitioned table in Hive的後續操作。我試圖在答案中使用建議,但無法使其在Spark 1.6.1中工作。Spark分區:創建RDD分區,但不創建Hive分區

我試圖從`DataFrame以編程方式創建分區。下面是相關代碼(改編自火花試驗):

hc.setConf("hive.metastore.warehouse.dir", "tmp/tests") 
// hc.setConf("hive.exec.dynamic.partition", "true") 
// hc.setConf("hive.exec.dynamic.partition.mode", "nonstrict") 
hc.sql("create database if not exists tmp") 
hc.sql("drop table if exists tmp.partitiontest1") 

Seq(2012 -> "a").toDF("year", "val") 
    .write 
    .partitionBy("year") 
    .mode(SaveMode.Append) 
    .saveAsTable("tmp.partitiontest1") 
hc.sql("show partitions tmp.partitiontest1").show 

完整的文件是在這裏:https://gist.github.com/SashaOv/7c65f03a51c7e8f9c9e018cd42aa4c4a

分區文件被創建優良的文件系統上,但蜂巢抱怨說,表未分區:

====================== 
HIVE FAILURE OUTPUT 
====================== 
SET hive.support.sql11.reserved.keywords=false 
SET hive.metastore.warehouse.dir=tmp/tests 
OK 
OK 
FAILED: Execution Error, return code 1 from org.apache.hadoop.hive.ql.exec.DDLTask. Table tmp.partitiontest1 is not a partitioned table 
====================== 

看起來像根本原因是org.apache.spark.sql.hive.HiveMetastoreCatalog.newSparkSQLSpecificMetastoreTable始終創建帶有空分區的表。

任何有助於推動這一進步的讚賞。

編輯:也創造SPARK-14927

回答

1

我發現了一個解決辦法:用它,如果你預先創建表,然後saveAsTable()就不會亂。所以下面的工作:

hc.setConf("hive.metastore.warehouse.dir", "tmp/tests") 
// hc.setConf("hive.exec.dynamic.partition", "true") 
// hc.setConf("hive.exec.dynamic.partition.mode", "nonstrict") 
hc.sql("create database if not exists tmp") 
hc.sql("drop table if exists tmp.partitiontest1") 

// Added line: 
hc.sql("create table tmp.partitiontest1(val string) partitioned by (year int)") 


Seq(2012 -> "a").toDF("year", "val") 
    .write 
    .partitionBy("year") 
    .mode(SaveMode.Append) 
    .saveAsTable("tmp.partitiontest1") 
hc.sql("show partitions tmp.partitiontest1").show 

此解決方案在1.6.1工作,但不是在1.5.1