Thanks for the blocking bug definitions, Luis.  They seem reasonable to me.

What about SRs though?  Should they be treated differently?  I was under the
impression that, in general, only bug fixes would go in to SRs.  If that's
the case then we can hold SRs to a slightly higher standard.

BTW, for this specific issue and the -li plugin wouldn't we expect forward
thinking consumers to be using it, assuming they realize this is the plugin
that will be default going forward.  I do remember talking with someone
at our last Summit (9-10mo ago) and they implied that real customers were
using the -li plugin version because it was more stable.  Are we breaking
them with Be SR2?

JamO

On 05/03/2016 11:58 PM, Luis Gomez wrote:
> Hi Jamo, I agree with you, we should avoid any kind of regression when we 
> release something.
> 
> Regarding what is blocker we have this definition for reference in the 
> Weather wiki:
> 
> 
>       Blocking Bugs definition
> 
> Blocking bugs may meet one or more of the following conditions:
> 
>   * Cause a major controller-wide regression. e.g: A bug in openflowplugin 
> prevents OpenFlow switches from connecting.
>   * Cause a major regression in another project. e.g: A bug in OVSDB causes 
> GBP service to fail.
>   * Dramatically decrease controller performance or scale. e.g: A bug in 
> yangtools causes OpenFlow perf test results to decrease by an order of 
> magnitude.
>   * Are immediately obvious to users. e.g: A bug in DLUX makes topology 
> unavailable in the GUI.
> 
> Generally, major regressions are likely candidates. Of course, terms like 
> "major" require substantial case-by-case interpretation.
> 
> 
> This particular issue: "OF plugin Li feature fails to deploy in cluster" 
> would be a "major" controller regression if it was not because afaik and
> after lot of discussion we never released the Li plugin in Beryllium, not 
> even as individual feature. So to me this is a "major regression" in a non
> released functionality which makes the bug non blocking imo.
> 
> BR/Luis
> 
> 
>> On May 3, 2016, at 10:30 PM, Jamo Luhrsen <[email protected] 
>> <mailto:[email protected]>> wrote:
>>
>> I know I'm the oddball here, but I am against releasing an SR with a known 
>> regression.  Even
>> if the regression is in a non-default feature, or intermittent or some other 
>> rare corner of
>> our product.  If it's truly a regression, then something has changed that we 
>> should track
>> down and resolve.
>>
>> Hopefully that can be part of our culture and process.  If regressions are 
>> sneaking in, then
>> I say we have to pay the price to fix them and try to get better at 
>> preventing them.
>>
>> We have nowhere near 100% system test coverage (if that's even possible), so 
>> one worry is
>> that whatever this regression is lurking in other places.
>>
>> it's only my 2c.
>>
>> JamO
>>
>>
>>
>>
>> On 05/03/2016 06:18 PM, Abhijit Kumbhare wrote:
>>> Yes - that is what we were discussing on the OpenFlow Plugin IRC. This is 
>>> not a blocker since it does not affect the end users - however we can wait
>>> till Wednesday morning (say 10 am Pacific) to get a response/fix from Jozef 
>>> (with Michal's help in merging it). Apparently it takes 24 hours for a
>>> respin - 12 hours to build & 12 hours to get approvals from project leads.
>>> If the ETA for the fix in the morning from Jozef pushes the build beyond 
>>> the TSC meeting - then we should skip.
>>>
>>> On Tue, May 3, 2016 at 5:55 PM, An Ho <[email protected] 
>>> <mailto:[email protected]> <mailto:[email protected]>> wrote:
>>>
>>>    I agree with this Anil.  Beryllium end users should not be using the Li 
>>> Plugin Design.____
>>>
>>>    __ __
>>>
>>>    If Jamo, Abhijit, and Jozef agrees with this, then we should mark the 
>>> issue as not a regression blocker and OKAY to release.____
>>>
>>>    __ __
>>>
>>>    Best Regards,____
>>>
>>>    An Ho____
>>>
>>>    __ __
>>>
>>>    *From:*Anil Vishnoi [mailto:[email protected] 
>>> <mailto:[email protected]>]
>>>    *Sent:* Tuesday, May 03, 2016 5:48 PM
>>>    *To:* Abhijit Kumbhare
>>>    *Cc:* An Ho; [email protected] 
>>> <mailto:[email protected]>
>>> <mailto:[email protected]>
>>>
>>>
>>>    *Subject:* Re: [openflowplugin-dev] Li plugin cluster broken____
>>>
>>>    __ __
>>>
>>>    In my opinion this is not a blocker, given that this plugin is not a 
>>> default plugin for Beryllium release.____
>>>
>>>    __ __
>>>
>>>    On Tue, May 3, 2016 at 2:45 PM, Abhijit Kumbhare <[email protected] 
>>> <mailto:[email protected]> <mailto:[email protected]>> wrote:____
>>>
>>>    I would like to know Jozef's thoughts on this.____
>>>
>>>    __ __
>>>
>>>    On Tue, May 3, 2016 at 1:15 PM, An Ho <[email protected] 
>>> <mailto:[email protected]> <mailto:[email protected]>> wrote:____
>>>
>>>    Have we been able to determine if this is a blocker, in the sense that 
>>> some end-user functionality that worked in SR1 no longer works in SR2?
>>>    Also, do have a known workaround for the issue?  Could we release SR2 
>>> with this known bug and workaround and target a fix in SR3?
>>>
>>>    Another concern I have is that Beryllium SR2 Build 20160425 did not find 
>>> any regression issues in openflowplugin and we have not merged nay
>>>    patches that impacted openflowplugin in the Beryllium branches.
>>>
>>>    Best Regards,
>>>    An Ho
>>>
>>>    From: Abhijit Kumbhare abhijitkoss at gmail.com <http://gmail.com> 
>>> <http://gmail.com>
>>>    Subject: Tue May 3 15:49:42 UTC 2016
>>>
>>>
>>>    Can you share the list of the patches Jozef?
>>>
>>>    Thanks,
>>>    Abhijit
>>>
>>>    On Tue, May 3, 2016 at 8:47 AM, Jamo Luhrsen <jluhrsen at gmail.com 
>>> <http://gmail.com> <http://gmail.com>> wrote:
>>>
>>>> just to confirm that this should not be a blocker for SR2, right?
>>>>
>>>> if so, we'll need your patches merged (and associated with a bug) and a
>>>> respin of
>>>> SR2.
>>>>
>>>> Thanks,
>>>> JamO
>>>>
>>>> On 05/03/2016 01:56 AM, Jozef Bacigal -X (jbacigal - PANTHEON TECHNOLOGIES
>>>> at Cisco) wrote:
>>>>> This is happening, occasionally, we solved it already in plugin. Patches
>>>> are ready to be merged we just wait for beryllium to be unlocked.
>>>>>
>>>>> Jozef
>>>>>
>>>>> -----Original Message-----
>>>>> From: openflowplugin-dev-bounces at lists.opendaylight.org 
>>>>> <http://lists.opendaylight.org> <http://lists.opendaylight.org> [mailto:
>>>> openflowplugin-dev-bounces at lists.opendaylight.org 
>>>> <http://lists.opendaylight.org> <http://lists.opendaylight.org>] On Behalf 
>>>> Of Jamo
>>>> Luhrsen
>>>>> Sent: 3. mája 2016 5:57
>>>>> To: Luis Gomez <ecelgp at gmail.com <http://gmail.com> 
>>>>> <http://gmail.com>>; openflowplugin-dev <
>>>> openflowplugin-dev at lists.opendaylight.org 
>>>> <http://lists.opendaylight.org> <http://lists.opendaylight.org>>
>>>>> Subject: Re: [openflowplugin-dev] Li plugin cluster broken
>>>>>
>>>>> This looks like something that used to happen from time to time.  Oldest
>>>> example was back on 4/17.  However, it currently looks like it's happening
>>>> almost every time.  That's a bit worrisome.
>>>>>
>>>>> I tried to recreate locally, but could not.
>>>>>
>>>>> not sure if that's helpful or not, but wanted to share.
>>>>>
>>>>> JamO
>>>>>
>>>>> On 05/02/2016 10:19 AM, Luis Gomez wrote:
>>>>>> Hi,
>>>>>>
>>>>>> It seems Li plugin cluster is broken for few days now. The ERROR is
>>>>>> 401 Unauthorized answer when polling the NB REST API. This happens when
>>>> controller does not start properly.
>>>>>>
>>>>>> https://jenkins.opendaylight.org/releng/view/CSIT-3node/job/openflowpl
>>>>>> ugin-csit-3node-clustering-only-boron
>>>>>>
>>>>>> _https://jenkins.opendaylight.org/releng/view/CSIT-3node/job/openflowp
>>>>>> lugin-csit-3node-clustering-only-beryllium/_
>>>>>>
>>>>>> BR/Luis
>>>>>>
>>>>>>
>>>>>>
>>>>>> _______________________________________________ openflowplugin-dev
>>>>>> mailing list openflowplugin-dev at lists.opendaylight.org 
>>>>>> <http://lists.opendaylight.org>
>>>>>> https://lists.opendaylight.org/mailman/listinfo/openflowplugin-dev
>>>>>>
>>>>> _______________________________________________
>>>>> openflowplugin-dev mailing list
>>>>> openflowplugin-dev at lists.opendaylight.org 
>>>>> <http://lists.opendaylight.org> <http://lists.opendaylight.org>
>>>>> https://lists.opendaylight.org/mailman/listinfo/openflowplugin-dev
>>>>>
>>>> _______________________________________________
>>>> openflowplugin-dev mailing list
>>>> openflowplugin-dev at lists.opendaylight.org 
>>>> <http://lists.opendaylight.org> <http://lists.opendaylight.org>
>>>> https://lists.opendaylight.org/mailman/listinfo/openflowplugin-dev
>>>> ____
>>>
>>>    __ __
>>>
>>>
>>>    _______________________________________________
>>>    openflowplugin-dev mailing list
>>>    [email protected] 
>>> <mailto:[email protected]> 
>>> <mailto:[email protected]>
>>>    https://lists.opendaylight.org/mailman/listinfo/openflowplugin-dev____
>>>
>>>
>>>
>>>    ____
>>>
>>>    __ __
>>>
>>>    -- ____
>>>
>>>    Thanks____
>>>
>>>    Anil____
>>>
>>>
>> _______________________________________________
>> openflowplugin-dev mailing list
>> [email protected] 
>> <mailto:[email protected]>
>> https://lists.opendaylight.org/mailman/listinfo/openflowplugin-dev
> 
_______________________________________________
openflowplugin-dev mailing list
[email protected]
https://lists.opendaylight.org/mailman/listinfo/openflowplugin-dev
  • [openflowplug... An Ho
    • Re: [ope... Abhijit Kumbhare
      • Re: ... Anil Vishnoi
        • ... An Ho
          • ... Abhijit Kumbhare
            • ... Jamo Luhrsen
              • ... Luis Gomez
              • ... Jozef Bacigal -X (jbacigal - PANTHEON TECHNOLOGIES at Cisco)
              • ... Abhijit Kumbhare
              • ... Jozef Bacigal -X (jbacigal - PANTHEON TECHNOLOGIES at Cisco)
              • ... Jamo Luhrsen
              • ... Luis Gomez
              • ... Jamo Luhrsen
              • ... Luis Gomez
              • ... Jamo Luhrsen

Reply via email to