Guys, sorry for later answer, I was working on small changes to new patch that solves the problem. I agree that this is not a blocker but we should not have any regressions in SRs or other releases neither. This particular bug: we were not able to catch it in time when prepared the changes for cluster. I wish I could merge this actual patch before the SR2 lock, but we weren’t sure with the results and tested it too long. Unfortunately in day we were ready to merge it was already a release log. ☺
So if you wish you can check the patch here https://git.opendaylight.org/gerrit/#/c/38369/3 We are testing now all possible scenarios on cluster, we try to cover all possibility and we are adding also a lot of junit test to catch any small bug we can possibly made with changes. Jozef From: [email protected] [mailto:[email protected]] On Behalf Of Luis Gomez Sent: 4. mája 2016 8:59 To: Jamo Luhrsen <[email protected]> Cc: An Ho <[email protected]>; [email protected] Subject: Re: [openflowplugin-dev] Li plugin cluster broken 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
