hadoop-common-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Matt Foley <mfo...@hortonworks.com>
Subject Re: [VOTE] Hadoop-1.0.1 Release Candidate 1
Date Mon, 20 Feb 2012 07:52:32 GMT
Giri found an issue with the naming of the 64-bit RPM and DEB; see
HADOOP-8090 <https://issues.apache.org/jira/browse/HADOOP-8090>.  The RPM
is called "hadoop-1.0.1-1.amd64.rpm" but should be called
"hadoop-1.0.1-1.x86_64.rpm".  Even though it's been this way for a long
time, apparently it causes an issue when the rpm is used with yum.  (BigTop
already uses the "x86_64" form.)  To fix this requires a change to

In addition, Nathan has requested patches for the following bugs:

   - MAPREDUCE-3824
<https://issues.apache.org/jira/browse/MAPREDUCE-3824>- Distributed
caches are not removed properly
   - MAPREDUCE-3851 <https://issues.apache.org/jira/browse/MAPREDUCE-3851> -
   Allow more aggressive action in detection of Jetty issues
   - HADOOP-8050 <https://issues.apache.org/jira/browse/HADOOP-8050>-
Deadlock in Namenode metrices

and Owen feels this one is pretty important:

   - HADOOP-8043 <https://issues.apache.org/jira/browse/HADOOP-8043> -
   and friends have some problems

So altogether that's five new issues, all raised (or raised in visibility)
since the 1.0.1 vote was called.  Two of them, HADOOP-8050 and
MAPREDUCE-3824, already have approved fixes, but the remaining three are
still in discussion.

The thing is that there are people waiting on the 1.0.1 release with some
urgency, including a pending BigTop release.  So instead of pulling the
1.0.1 release, waiting for fixes, then waiting an additional week for
another vote, I suggest that we let 1.0.1 proceed as is, and do a 1.0.2
release as soon as the remaining three bugs noted above are fixed.

If you agree with this, please vote in favor of the 1.0.1 release.  If you
disagree, please vote against the release, and I'll do a 1.0.1-rc2 instead
of a 1.0.2.  As a reminder, the vote closes Tuesday afternoon.

The goal here is to keep everyone fully informed, while providing the most
good to the most people, in the face of conflicting requests.

On Tue, Feb 14, 2012 at 2:29 PM, Matt Foley <mattf@apache.org> wrote:

> Hi all,
> Hadoop-1.0.1-rc1 is now available at
>     http://people.apache.org/~mattf/hadoop-1.0.1-rc1/
> Please download it, test it, and vote on whether to accept it as Release
> 1.0.1.
> Voting will close at 2:30pm PST on Tuesday 21 February.
> Thank you,
> --Matt (Release Manager)
> This is a maintenance release from 1.0.0, but a number of
> production-relevant bug fixes were requested for inclusion in the last week
> or so.  The complete release notes are available at
> http://people.apache.org/~mattf/hadoop-1.0.1-rc1/RELEASE_NOTES_HADOOP-1.0.1.html
> The summary list is:
>    - HADOOP-8009 <https://issues.apache.org/jira/browse/HADOOP-8009>.
>    Create hadoop-client and hadoop-minicluster artifacts for downstream
>    projects
>    - HADOOP-8037 <https://issues.apache.org/jira/browse/HADOOP-8037>. Binary
>    tarball does not preserve platform info for native builds, and RPMs fail to
>    provide needed symlinks for libhadoop.so
>    - MAPREDUCE-3184 <https://issues.apache.org/jira/browse/MAPREDUCE-3184>. Improve
>    handling of fetch failures when a tasktracker is not responding on HTTP
>    - HADOOP-7470 <https://issues.apache.org/jira/browse/HADOOP-7470>.
>    move up to Jackson 1.8.8
>    - HADOOP-7960 <https://issues.apache.org/jira/browse/HADOOP-7960>. Port
>    HADOOP-5203 to branch-1, build version comparison is too restrictive
>    - HADOOP-7964 <https://issues.apache.org/jira/browse/HADOOP-7964>. Deadlock
>    in class init.
>    - HADOOP-7987 <https://issues.apache.org/jira/browse/HADOOP-7987>. Support
>    setting the run-as user in unsecure mode
>    - HADOOP-7988 <https://issues.apache.org/jira/browse/HADOOP-7988>. Upper
>    case in hostname part of the principals doesn't work with kerberos.
>    - HADOOP-8010 <https://issues.apache.org/jira/browse/HADOOP-8010>. hadoop-config.sh
>    spews error message when HADOOP_HOME_WARN_SUPPRESS is set to true and
>    HADOOP_HOME is present
>    - HADOOP-8052 <https://issues.apache.org/jira/browse/HADOOP-8052>. Hadoop
>    Metrics2 should emit Float.MAX_VALUE (instead of Double.MAX_VALUE) to avoid
>    making Ganglia's gmetad core
>    - HDFS-2379 <https://issues.apache.org/jira/browse/HDFS-2379>. 0.20:
>    Allow block reports to proceed without holding FSDataset lock
>    - HDFS-2814 <https://issues.apache.org/jira/browse/HDFS-2814>. NamenodeMXBean
>    does not account for svn revision in the version information
>    - MAPREDUCE-3343 <https://issues.apache.org/jira/browse/MAPREDUCE-3343>. TaskTracker
>    Out of Memory because of distributed cache
>    - MAPREDUCE-3607 <https://issues.apache.org/jira/browse/MAPREDUCE-3607>. Port
>    missing new API mapreduce lib classes to 1.x

  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message