Hi all,
Release Candidate for Hadoop-1.0.2 is now available for vote, at
    http://people.apache.org/~mattf/hadoop-1.0.2-rc1/

There are 19 bug fixes and enhancements in this release, including:

HADOOP-7206. added Snappy compression
HDFS-3075. automatically restore previously failed storage directories
MAPREDUCE-3773. Add queue metrics for bucketized job run times
Please see the Release Notes for the full list, at:
    http://people.apache.org/~mattf/hadoop-1.0.2-rc1/releasenotes.html

Vote will run the statutory 7 days, and close on Monday 26 Mar at 11:30am
PDT.

Thank you,
--Matt
Release Manager, Hadoop 1.0

I have to withdraw RC-1, due to a blocker bug found:
HADOOP-8201 breaks the Snappy compression/decompression

I will regenerate it tonight.
Thanks,
--Matt
On Mon, Mar 19, 2012 at 11:21 AM, Matt Foley <[EMAIL PROTECTED]> wrote:

> Hi all,
> Release Candidate for Hadoop-1.0.2 is now available for vote, at
>     http://people.apache.org/~mattf/hadoop-1.0.2-rc1/
>
> There are 19 bug fixes and enhancements in this release, including:
>
> HADOOP-7206. added Snappy compression
> HDFS-3075. automatically restore previously failed storage directories
> MAPREDUCE-3773. Add queue metrics for bucketized job run times
>
>
> Please see the Release Notes for the full list, at:
>     http://people.apache.org/~mattf/hadoop-1.0.2-rc1/releasenotes.html
>
> Vote will run the statutory 7 days, and close on Monday 26 Mar at 11:30am
> PDT.
>
> Thank you,
> --Matt
> Release Manager, Hadoop 1.0
>

Hi Matt,

I don't understand why this issue is considered a blocker. The JIRA is marked (rightly, I think) as an improvement, not a bug.

Can you perhaps explain further why this warrants a new RC?

Thanks a lot,
Aaron

On Mar 22, 2012, at 7:24 PM, Matt Foley <[EMAIL PROTECTED]> wrote:

> I have to withdraw RC-1, due to a blocker bug found:
> HADOOP-8201 breaks the Snappy compression/decompression
>
> I will regenerate it tonight.
> Thanks,
> --Matt
>
>
> On Mon, Mar 19, 2012 at 11:21 AM, Matt Foley <[EMAIL PROTECTED]> wrote:
>
>> Hi all,
>> Release Candidate for Hadoop-1.0.2 is now available for vote, at
>>    http://people.apache.org/~mattf/hadoop-1.0.2-rc1/
>>
>> There are 19 bug fixes and enhancements in this release, including:
>>
>> HADOOP-7206. added Snappy compression
>> HDFS-3075. automatically restore previously failed storage directories
>> MAPREDUCE-3773. Add queue metrics for bucketized job run times
>>
>>
>> Please see the Release Notes for the full list, at:
>>    http://people.apache.org/~mattf/hadoop-1.0.2-rc1/releasenotes.html
>>
>> Vote will run the statutory 7 days, and close on Monday 26 Mar at 11:30am
>> PDT.
>>
>> Thank you,
>> --Matt
>> Release Manager, Hadoop 1.0
>>

Sorry for lack of clarity; I have now commented in the jira.  I believe
this bug caused Snappy to not be built correctly, which in turn causes MR
jobs to fail if Snappy compression is configured on.
--Matt

On Thu, Mar 22, 2012 at 7:40 PM, Aaron T. Myers <[EMAIL PROTECTED]> wrote:

> Hi Matt,
>
> I don't understand why this issue is considered a blocker. The JIRA is
> marked (rightly, I think) as an improvement, not a bug.
>
> Can you perhaps explain further why this warrants a new RC?
>
> Thanks a lot,
> Aaron
>
> On Mar 22, 2012, at 7:24 PM, Matt Foley <[EMAIL PROTECTED]> wrote:
>
> > I have to withdraw RC-1, due to a blocker bug found:
> > HADOOP-8201 breaks the Snappy compression/decompression
> >
> > I will regenerate it tonight.
> > Thanks,
> > --Matt
> >
> >
> > On Mon, Mar 19, 2012 at 11:21 AM, Matt Foley <[EMAIL PROTECTED]> wrote:
> >
> >> Hi all,
> >> Release Candidate for Hadoop-1.0.2 is now available for vote, at
> >>    http://people.apache.org/~mattf/hadoop-1.0.2-rc1/
> >>
> >> There are 19 bug fixes and enhancements in this release, including:
> >>
> >> HADOOP-7206. added Snappy compression
> >> HDFS-3075. automatically restore previously failed storage directories
> >> MAPREDUCE-3773. Add queue metrics for bucketized job run times
> >>
> >>
> >> Please see the Release Notes for the full list, at:
> >>    http://people.apache.org/~mattf/hadoop-1.0.2-rc1/releasenotes.html
> >>
> >> Vote will run the statutory 7 days, and close on Monday 26 Mar at
> 11:30am
> >> PDT.
> >>
> >> Thank you,
> >> --Matt
> >> Release Manager, Hadoop 1.0
> >>
>

On Thu, Mar 22, 2012 at 11:12 PM, Matt Foley <[EMAIL PROTECTED]> wrote:

> I believe
> this bug caused Snappy to not be built correctly, which in turn causes MR
> jobs to fail if Snappy compression is configured on.
>

Ah, I see. Thanks for the clarification.

Given that this fix is quite small, and given that no one has so far voted
on RC-1, I'd personally be in favor of not restarting the voting period for
RC-2. Thoughts?

--
Aaron T. Myers
Software Engineer, Cloudera

On Thu, Mar 22, 2012 at 11:12 PM, Matt Foley <[EMAIL PROTECTED]> wrote:
> Sorry for lack of clarity; I have now commented in the jira.  I believe
> this bug caused Snappy to not be built correctly, which in turn causes MR
> jobs to fail if Snappy compression is configured on.

Not for this release, obviously, but in general perhaps the right solution
here is unbundle native snappy the same way jsvc got unbundled.

Thanks,
Roman.