[ https://issues.apache.org/jira/browse/SPARK-11214?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15087675#comment-15087675 ]
Reynold Xin edited comment on SPARK-11214 at 1/8/16 11:29 PM: -------------------------------------------------------------- Guys, I have a reason for the joining Problem. Use this code in pyspark: {code} from pyspark import SparkContext, SparkConf from pyspark.sql import HiveContext conf = (SparkConf() .setAppName("joining_test") .setMaster("yarn-client") .set("spark.driver.cores", "20") .set("spark.driver.memory", "50g") .set("spark.driver.maxResultSize", "0") .set("spark.num.executors", "5") .set("spark.executor.cores", "10") # .set("spark.executor.memory", "40G") ) sc = SparkContext(conf=conf) hiveContext = HiveContext(sc) hiveContext.sql('SELECT * FROM (SELECT "a" AS a) AS a JOIN (SELECT "a" AS b) AS b ON a.a = b.b').take(10) {code} When ".set("spark.executor.memory", "40G")" is commented out the result will be right. When the executor momory is set the result will be empty. Could somebody validate this? Thank you was (Author: xdsticker): Guys, I have a reason for the joining Problem. Use this code in pyspark: from pyspark import SparkContext, SparkConf from pyspark.sql import HiveContext conf = (SparkConf() .setAppName("joining_test") .setMaster("yarn-client") .set("spark.driver.cores", "20") .set("spark.driver.memory", "50g") .set("spark.driver.maxResultSize", "0") .set("spark.num.executors", "5") .set("spark.executor.cores", "10") # .set("spark.executor.memory", "40G") ) sc = SparkContext(conf=conf) hiveContext = HiveContext(sc) hiveContext.sql('SELECT * FROM (SELECT "a" AS a) AS a JOIN (SELECT "a" AS b) AS b ON a.a = b.b').take(10) When ".set("spark.executor.memory", "40G")" is commented out the result will be right. When the executor momory is set the result will be empty. Could somebody validate this? Thank you > Join with Unicode-String results wrong empty > -------------------------------------------- > > Key: SPARK-11214 > URL: https://issues.apache.org/jira/browse/SPARK-11214 > Project: Spark > Issue Type: Bug > Components: PySpark, SQL > Affects Versions: 1.5.0 > Reporter: Hans Fischer > Assignee: Josh Rosen > Priority: Critical > Fix For: 1.5.1 > > > I created a join that should clearly result in a single row but return: > empty. Could someone validate this bug? > hiveContext.sql('SELECT * FROM (SELECT "c" AS a) AS a JOIN (SELECT "c" AS b) > AS b ON a.a = b.b').take(10) > result: [] > kind regards > Hans -- This message was sent by Atlassian JIRA (v6.3.4#6332) --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org For additional commands, e-mail: issues-h...@spark.apache.org