[ https://issues.apache.org/jira/browse/MAHOUT-2023?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16282454#comment-16282454 ]
ASF GitHub Bot commented on MAHOUT-2023: ---------------------------------------- Github user andrewpalumbo commented on a diff in the pull request: https://github.com/apache/mahout/pull/353#discussion_r155627595 --- Diff: spark/src/main/assembly/dependency-reduced.xml --- @@ -34,16 +34,15 @@ </unpackOptions> <scope>runtime</scope> <outputDirectory>/</outputDirectory> - <useTransitiveFiltering>true</useTransitiveFiltering> + <useProjectArtifact>true</useProjectArtifact> <includes> <!-- guava only included to get Preconditions in mahout-math and mahout-hdfs --> <include>com.google.guava:guava</include> - <include>com.github.scopt_${scala.compat.version}</include> + <include>com.github.scopt:scopt_2.10</include> + <include>com.github.scopt:scopt_2.11</include> <include>com.tdunning:t-digest</include> <include>org.apache.commons:commons-math3</include> <include>it.unimi.dsi:fastutil</include> - <include>org.apache.mahout:mahout-native-viennacl_${scala.compat.version}</include> --- End diff -- I think after we discussed yesterday, if adding these jars to the classpath is the better practice, and we want this jar to be a slim as possible, leaving them out would be fine. I am unclear on how often its shipped around on the backend. I think that the spark method- just adding to the `spark.executor.extraClassPath` and `spark.driver.extraClassPath` makes sense. @rawkintrevo this is what you were saying we should to, correct and to just add it to the documentation? I don't think that the `dependency-reduced.jar` is the way we want to be shipping everything in the future. I think that this is something that we as a project need to discuss at some point. The original goal for these solver modules (and the somewhat current, hacky implementation) was to probe the classpath for the existence of a solver, and then to use e.g. viennacl-omp#Mmul if it is on the classpath. Not sure that adding it to the dependency reduce jar is the way to go. I say we just ship this as is- still need to do more testing on a cluster due to pat's next scopt find: ``` Exception in thread "main" java.lang.NoClassDefFoundError: scala/Product at java.lang.ClassLoader.defineClass1(Native Method) at java.lang.ClassLoader.defineClass(ClassLoader.java:763) ``` > Drivers broken, scopt classes not found > --------------------------------------- > > Key: MAHOUT-2023 > URL: https://issues.apache.org/jira/browse/MAHOUT-2023 > Project: Mahout > Issue Type: Bug > Components: build > Affects Versions: 0.13.1 > Environment: any > Reporter: Pat Ferrel > Assignee: Trevor Grant > Priority: Blocker > Fix For: 0.13.1 > > > Type `mahout spark-itemsimilarity` after Mahout is installed properly and you > get a fatal exception due to missing scopt classes. > Probably a build issue related to incorrect versions of scopt being looked > for. -- This message was sent by Atlassian JIRA (v6.4.14#64029)