hadoop-common-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Martin Walsh <martin.wa...@oracle.com>
Subject Re: JIRA dependencies and Jenkins
Date Wed, 22 Jul 2015 15:02:48 GMT
On 07/22/15 03:08 PM, Chris Nauroth wrote:
> Jenkins will not recognize inter-dependencies of multiple jiras and try to
> pull in dependent patches.  You'd need to wait for the dependent patch to
> get reviewed and committed first, and then trigger the Jenkins run for the
> second patch.

OK.

>
> Sometimes people like to get a "preview" Jenkins run to see an early
> health check instead of waiting for the commit.  You can do that by
> uploading a consolidated patch file to the second jira that includes all
> of the changes and applies cleanly to trunk.  After the first patch gets
> committed, you would once again upload a new patch file on the second jira
> containing just the second set of changes.  (The final Jenkins run should
> be done on what is actually getting committed within the scope of that
> second jira.)

Looks like I did a preview Jenkins run by mistake.  OK, thanks for the 
info.  Will wait for the integration of the first patch.

Martin

>
> Thanks for the patches, Martin!
>
> --Chris Nauroth
>
>
>
>
> On 7/22/15, 6:29 AM, "Martin Walsh" <martin.walsh@oracle.com> wrote:
>
>> I am working on the following two bugs:
>>
>> https://issues.apache.org/jira/browse/HADOOP-12184
>> https://issues.apache.org/jira/browse/HADOOP-7824
>>
>> HADOOP-12184 is in the "Patch Available" state and is marked as a
>> requirement for HADOOP-7824.  When I submit the patch for HADOOP-7824 it
>> will touch the same section of code that HADOOP-12184 has done.
>> Therefore my patch will only apply cleanly if HADOOP-12184 patch is
>> applied first.  Is Jenkins smart enough to pull in required patches as
>> detailed in the JIRA and apply them first, or does the dependency need
>> to be committed back to the repository?
>>
>> Thanks,
>>
>> Martin
>>
>


Mime
View raw message