The concern that I have here is that NIFI-938 and NIFI-939 appear to be 
separate issues.
The patch for NIFI-939 should fix the symptoms described by 938. I am 
concerned, though,
that it is simply avoiding the condition that caused 938. I.e., NIFI-938 causes 
the bug described
in NIFI-939 not to be exposed. I want to do some more investigating
to try to figure out what is causing the issues that Brian saw, rather than 
accepting the patch
in 939 as a fix.

----------------------------------------
> From: [email protected]
> Date: Wed, 9 Sep 2015 12:05:44 +0000
> Subject: Re: 0.3.0 release
> To: [email protected]
>
> Brian,
>
> In looking at the ticket, it sounds like you're almost satisfied with the
> fix. Can you let us know when you're convinced it's working? Thanks.
>
> Brandon
>
> On Tue, Sep 8, 2015 at 3:51 PM Brian Ghigiarelli <[email protected]>
> wrote:
>
>> Hi Brandon,
>>
>> With the latest 0.3.0-SNAPSHOT on master, I've started seeing an issue
>> appending content to FlowFiles (not reproducible with unit tests, but is
>> reproducible by running the GetKafka processor when it is configured for
>> batching).
>>
>> Because a big part of the Content Repo was reworked with this release, I'd
>> ask that you consider a ticket I just created:
>> https://issues.apache.org/jira/browse/NIFI-938
>>
>> Thanks,
>> Brian
>>
>> On Tue, Sep 8, 2015 at 3:05 PM, Brandon DeVries <[email protected]> wrote:
>>
>>> All,
>>>
>>> It looks like there's only one ticket (NIFI-936[1]) unresolved that is
>>> slated for 0.3.0[2]. Are there any other issues that anyone believes
>> need
>>> to be in the 0.3.0 release? If not, does anyone object to shooting for
>> an
>>> 0.3.0 release shortly after NIFI-936 is resolved?
>>>
>>> [1] https://issues.apache.org/jira/browse/NIFI-936
>>> [2]
>>>
>>>
>> https://issues.apache.org/jira/issues/?filter=12331975&jql=project%20%3D%20NIFI%20AND%20fixVersion%20%3D%200.3.0%20%20order%20by%20status%20asc
>>>
>>> Brandon
>>
                                          

Reply via email to