release tar.gz contains duplicate files

• Key: HADOOP-3477
• Type: Bug
• Status: Resolved
• Resolution: Fixed
• Priority: Major
• Assignee: Adam Heath
• Reporter: Adam Heath
• Component/s: build
• Affects Version/s: 0.16.4, 0.17.0
• Fix Version/s: 0.17.1

Comments

[Adam Heath]

Simple patch, one line.

[Adam Heath]

More details:

==
adam@zoot:~/code/upstream/nutch$ tar tzvf hadoop-0.17.0.tar.gz |awk '{print $6}'|sort|uniq -c|grep -v '1 '
2 hadoop-0.17.0/contrib/hod/bin/VERSION
2 hadoop-0.17.0/contrib/hod/bin/checknodes
2 hadoop-0.17.0/contrib/hod/bin/hod
2 hadoop-0.17.0/contrib/hod/bin/hodcleanup
2 hadoop-0.17.0/contrib/hod/bin/hodring
2 hadoop-0.17.0/contrib/hod/bin/ringmaster
==

This occurs because those files are included twice in build.xml, because there is no explicit exclude on those dirs, like there is for the others. The attached patch(1-line) adds such an exclude.

This is keeping me from using the downloaded tarball as the orig.tar.gz for a debian package.

[Doug Cutting]

I just committed this. Thanks, Adam!

[Adam Heath]

Could you apply the fix to 0.16 as well(creating 0.16.5)? Nutch isn't using 0.17 yet.

[Doug Cutting]

I guess this is a regression, since it was introduced in 0.16.0 by HADOOP-2494. Unless there are objections, I will merge it into the 0.16 and 0.17 branches.

[Hudson]

Integrated in Hadoop-trunk #511 (See http://hudson.zones.apache.org/hudson/job/Hadoop-trunk/511/)

[Hudson]

Integrated in Hadoop-trunk #523 (See http://hudson.zones.apache.org/hudson/job/Hadoop-trunk/523/)