Re: Error building spark app with Maven

2016-03-15 Thread Mich Talebzadeh
that should read anything.sbt Dr Mich Talebzadeh LinkedIn * https://www.linkedin.com/profile/view?id=AAEWh2gBxianrbJd6zP6AcPCCdOABUrV8Pw * http://talebzadehmich.wordpress.com On 16 March 2016 at 00:04,

Re: Error building spark app with Maven

2016-03-15 Thread Mich Talebzadeh
in mvn the build mvn package will look for a file called pom.xml in sbt the build sbt package will look for a file called anything.smt It works Keep it simple I will write a ksh script that will create both generic and sbt files on the fly in the correct directory (at the top of the tree) and

Re: Error building spark app with Maven

2016-03-15 Thread Jakob Odersky
The artifactId in maven basically (in a simple case) corresponds to name in sbt. Note however that you will manually need to append the _scalaBinaryVersion to the artifactId in case you would like to build against multiple scala versions (otherwise maven will overwrite the generated jar with the

Re: Error building spark app with Maven

2016-03-15 Thread Ted Yu
Feel free to adjust artifact Id and version in maven. They're under your control. > On Mar 15, 2016, at 4:27 PM, Mich Talebzadeh > wrote: > > ok Ted > > In sbt I have > > name := "ImportCSV" > version := "1.0" > scalaVersion := "2.10.4" > > which ends up in

Re: Error building spark app with Maven

2016-03-15 Thread Mich Talebzadeh
ok Ted In sbt I have name := "ImportCSV" version := "1.0" scalaVersion := "2.10.4" which ends up in importcsv_2.10-1.0.jar as part of *target/scala-2.10/importcsv_2.**10-1.0.jar* In mvn I have 1.0 scala Does it matter? Dr Mich Talebzadeh LinkedIn *

Re: Error building spark app with Maven

2016-03-15 Thread Ted Yu
1.0 ... scala On Tue, Mar 15, 2016 at 4:14 PM, Mich Talebzadeh wrote: > An observation > > Once compiled with MVN the job submit works as follows: > > + /usr/lib/spark-1.5.2-bin-hadoop2.6/bin/spark-submit --packages > com.databricks:spark-csv_2.11:1.3.0 --class

Re: Error building spark app with Maven

2016-03-15 Thread Mich Talebzadeh
An observation Once compiled with MVN the job submit works as follows: + /usr/lib/spark-1.5.2-bin-hadoop2.6/bin/spark-submit --packages com.databricks:spark-csv_2.11:1.3.0 --class ImportCSV --master spark:// 50.140.197.217:7077 --executor-memory=12G --executor-cores=12 --num-executors=2

Re: Error building spark app with Maven

2016-03-15 Thread Mich Talebzadeh
Many thanks Ted and thanks for heads up Jakob Just these two changes to dependencies org.apache.spark spark-core*_2.10* 1.5.1 org.apache.spark spark-sql*_2.10* 1.5.1 [DEBUG] endProcessChildren: artifact=spark:scala:jar:1.0 [INFO]

Re: Error building spark app with Maven

2016-03-15 Thread Jakob Odersky
Hi Mich, probably unrelated to the current error you're seeing, however the following dependencies will bite you later: spark-hive_2.10 spark-csv_2.11 the problem here is that you're using libraries built for different Scala binary versions (the numbers after the underscore). The simple fix here

Error building spark app with Maven

2016-03-15 Thread Mich Talebzadeh
Hi, I normally use sbt and using this sbt file works fine for me cat ImportCSV.sbt name := "ImportCSV" version := "1.0" scalaVersion := "2.10.4" libraryDependencies += "org.apache.spark" %% "spark-core" % "1.5.1" libraryDependencies += "org.apache.spark" %% "spark-sql" % "1.5.1"