Polina Marasanova created ZEPPELIN-926:
------------------------------------------
Summary: Getting 'File name too long' error when running scala
code in Zeppelin
Key: ZEPPELIN-926
URL: https://issues.apache.org/jira/browse/ZEPPELIN-926
Project: Zeppelin
Issue Type: Bug
Components: Interpreters
Affects Versions: 0.5.6
Reporter: Polina Marasanova
Priority: Minor
1. Create %spark paragraph
2. Run this code several times in a row (usually 10~15 is enough)
import org.apache.spark.sql.expressions.Window._
import org.apache.spark.sql.DataFrame
import org.apache.spark.sql.functions._
def normaliseDate(table: DataFrame, col: String) =
from_unixtime(table(col)/1000, "yyyy-MM-dd") as col
def normaliseTime(table: DataFrame, col: String) =
from_unixtime(table(col)/1000, "HH:mm:ss") as col
val df = sqlContext.sql("select * from parquet.`big_parquet_file`")
val result =
df.withColumn("start_date",from_unixtime((df("start_date")/1000),"yyyy-MM-dd"))
Result: Getting 'File name too long'
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
