You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@parquet.apache.org by "Ryan Blue (JIRA)" <ji...@apache.org> on 2015/04/10 18:19:12 UTC
[jira] [Commented] (PARQUET-249) Parquet does not compile with
-Phadoop-2
[ https://issues.apache.org/jira/browse/PARQUET-249?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14489875#comment-14489875 ]
Ryan Blue commented on PARQUET-249:
-----------------------------------
Hi Tony,
We don't actually use the hadoop-2 profile for compilation, just for
testing. We compile against hadoop-1 and use shims to run with any
Hadoop version. The hadoop-2 profile is how we verify that all of the
project tests work when running with hadoop-2.
As a side note, I think we should start a discussion at some point after
1.7.0 about when we're going to move off of Hadoop 1, and when we'll
start using Java 7 source.
rb
--
Ryan Blue
Software Engineer
Cloudera, Inc.
> Parquet does not compile with -Phadoop-2
> ----------------------------------------
>
> Key: PARQUET-249
> URL: https://issues.apache.org/jira/browse/PARQUET-249
> Project: Parquet
> Issue Type: Bug
> Components: parquet-mr
> Affects Versions: 1.6.0
> Environment: RHEL 7.1 on x86_64 & PPC64LE
> Reporter: Tony Reix
> Priority: Critical
>
> When running:
> mvn clean install -l ../mvn.CleanInstallTest.res -fn ; mvn test -l ../mvn.Test.res6 -fn -X
> I get 510 tests run and 0 error and 0 failure.
> If now I do (adding -Phadoop-2) :
> mvn clean install -l ../mvn.CleanInstallTest.res -fn -Phadoop-2
> Parquet no more compiles. With errors:
> [INFO] --- maven-compiler-plugin:3.1:testCompile (default-testCompile) @ parquet-common ---
> [INFO] Changes detected - recompiling the module!
> [INFO] Compiling 3 source files to /home/reixt/PARQUET/APACHE/incubator-parquet-mr/parquet-common/target/test-classes
> [INFO] -------------------------------------------------------------
> [ERROR] COMPILATION ERROR :
> [INFO] -------------------------------------------------------------
> [ERROR] /home/reixt/PARQUET/APACHE/incubator-parquet-mr/parquet-common/src/test/java/parquet/bytes/TestBytesUtil.java:[22,28] cannot find symbol
> symbol: class BytesUtils
> location: package parquet.bytes
> [ERROR] /home/reixt/PARQUET/APACHE/incubator-parquet-mr/parquet-common/src/test/java/parquet/bytes/TestBytesUtil.java:[22,1] static import only from classes and interfaces
> [ERROR] /home/reixt/PARQUET/APACHE/incubator-parquet-mr/parquet-common/src/test/java/parquet/bytes/TestBytesUtil.java:[30,21] cannot find symbol
> symbol: method getWidthFromMaxInt(int)
> Am I wrong ? or is there an issue with Parquet 1.6.0 ?
> I did:
> $ git clone https://github.com/apache/incubator-parquet-mr.git
> $ git status
> # On branch master
> $ mvn clean install -l ../mvn.CleanInstallTest.res-Phadoop-2 -fn -Phadoop-2
> No changes in source code.
> I've recreated my environment on x86_64 to check and to be sure. The same.
> Does it mean that Parquet is not compatible with Hadoop 2 ?
> Or an issue in pom.xml code ?
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)