Duplicate upgrade scripts issue

[Yedidyah Bar David]

On Tue, Jun 7, 2016 at 12:45 PM, Martin Perina <mperina at redhat.com> wrote:
>
>
> On Tue, Jun 7, 2016 at 11:42 AM, Eli Mesika <emesika at redhat.com> wrote:
>>
>> Ha , one more thing :
>>
>> We would like to force the existence of such hooks if possible ...
>
>
> +1
>
> The hook should be included by default after git clone if possible. And if
> it's then I'd also force inclusion of commit message hook which generates
> change-id
>

IMO having it as a git hook is a useful optimization, but more important
is to make it run on the server, to prevent push/merge.

To do this we can simply patch the engine's automation/check-patch.sh and/or
check-merged.sh .

>>
>> On Tue, Jun 7, 2016 at 12:41 PM, Eli Mesika <emesika at redhat.com> wrote:
>>>
>>> Hi guys
>>>
>>> I have talked with Eyal about the $Subject and he asked me to write and
>>> send this email
>>>
>>> As you probably know, we have from time to time an issue with duplicate
>>> upgrade scripts that are merged by mistake, each such issue forces us to
>>> publish a fixing patch that renames the duplicated file.
>>> I was discussed this issue today with Marin P on out weekly meeting
>>>
>>> We would like to write some kind of a hook that will check on each patch
>>> set if it has DB upgrade files and rename them (if necessary) such that it
>>> will have the correct numbering according to the last existing upgrade patch
>>> on the related branch.
>>>
>>> The hook should be done upon 'git push' request so it will also prevent
>>> CI tests to fail on this issue
>>>
>>> I will be happy to get your ideas/comments on that
>>>
>>> Thanks
>>>
>>> Eli Mesika
>>
>>
>
>
> _______________________________________________
> Infra mailing list
> Infra at ovirt.org
> http://lists.ovirt.org/mailman/listinfo/infra
>



-- 
Didi