Hi, Looks like the merge job is doing the wrong thing.
its checking out the patchset, rather than what is in master. since the patchset is from Dec 28, and does not touch ./docs/requirements/05-implementation.rst (which has been updated since then) git does not see a conflict with merging this code. If you look at the console output for the merge job https://build.opnfv.org/ci/job/opnfv-docs-merge-master/1232/console its doing this. > git fetch --tags --progress ssh://gerrit.opnfv.org:29418/$GERRIT_PROJECT refs/changes/65/24365/4 # timeout=15 and it should be just doing a straight clone of whats in master (after 24365 is merged) For now you can just submit a small change to doctor/docs/whatever and then merge it, this will get the latest version of all rst files into artifacts. I will submit at change to jjb/opnfvdocs/opnfvdocs.yml to fix this confusing behavior in the future. Regards, Aric On Thu, Jan 26, 2017 at 7:50 AM, Kunzmann, Gerald via RT <opnfv-helpd...@rt.linuxfoundation.org> wrote: > Dear all, > > Sorry for another email on this issue: > > Today the situation is even more strange: > > There is the old alarm table in generated artifacts file: > http://artifacts.opnfv.org/doctor/docs/requirements/index.html > But the git tree still contains the new table: > https://git.opnfv.org/doctor/tree/docs/requirements/05-implementation.rst > > What I realized is that in https://gerrit.opnfv.org/gerrit/#/c/24365/ after > Ryota had merged the patch, > jenkins-ci has just sent the document links, but there was no specific build > started. Is this normal/okay? > > Best regards, > Gerald > > -----Original Message----- > From: Kunzmann, Gerald > Sent: Donnerstag, 26. Januar 2017 13:40 > To: 'opnfv-helpd...@rt.linuxfoundation.org' > <opnfv-helpd...@rt.linuxfoundation.org>; 'Fatih Degirmenci' > <fatih.degirme...@ericsson.com> > Cc: 'opnfv-tech-discuss@lists.opnfv.org' > <opnfv-tech-discuss@lists.opnfv.org>; 'r-m...@cq.jp.nec.com' > <r-m...@cq.jp.nec.com> > Subject: RE: [OPNFV Helpdesk #35878] [linuxfoundation.org #35878] Re: > [opnfv-tech-discuss] [releng][helpdesk] Problem with Git/Gerrit/Jenkins > > Dear all, > > So this issue is easily reproducible. It had again happened today with > another patch being merged.... > > https://gerrit.opnfv.org/gerrit/24365 > > We are now back to the old table again :( > > We have to find a solution to this problem!!! > > Best regards, > Gerald > > -----Original Message----- > From: Kunzmann, Gerald > Sent: Mittwoch, 25. Januar 2017 13:56 > To: 'opnfv-helpd...@rt.linuxfoundation.org' > <opnfv-helpd...@rt.linuxfoundation.org> > Cc: opnfv-tech-discuss@lists.opnfv.org; r-m...@cq.jp.nec.com > Subject: [OPNFV Helpdesk #35878] [linuxfoundation.org #35878] Re: > [opnfv-tech-discuss] [releng][helpdesk] Problem with Git/Gerrit/Jenkins > > HI Fatih, > > Thanks for your quick check. The remerged had helped to bring back the > table... > > Let's see docs team's reply on it. > > I agree gerrit was fine not raising conflict as the changed line in the > "editorial" patch was not conflicting with the other patch. > Still, shouldn't gerrit have just changed that one line instead of also other > parts of the file and/or shouldn't gerrit have raised some warning it was > changing more than just the one line? > > When looking at the diff in gerrit for the "editorial" patch, it does not > compare against the latest master, otherwise it would have shown more changes > than just this one line. This might have caused the issue?!? > > Best regards, > Gerald > > -----Original Message----- > From: fatih.degirme...@ericsson.com via RT > [mailto:opnfv-helpd...@rt.linuxfoundation.org] > Sent: Mittwoch, 25. Januar 2017 13:38 > To: Kunzmann, Gerald <kunzm...@docomolab-euro.com> > Cc: opnfv-tech-discuss@lists.opnfv.org; r-m...@cq.jp.nec.com > Subject: [OPNFV Helpdesk #35878] [linuxfoundation.org #35878] Re: > [opnfv-tech-discuss] [releng][helpdesk] Problem with Git/Gerrit/Jenkins > > Hi, > > I had a quick look at this and diffed the commits and I see no conflict so > Gerrit was fine not raising conflict. (the changed line in > 05-implementation.rst in most recent patch is not a conflicting change) > > I then tried regenerating the documents by putting remerge as comment to > check the doc generation script/job output for possible errors which I didn't > see any. But the generated document is correct now. If you don't see the > updates, please try using a different browser or refresh the page after > couple of minutes due to caching. > > With that said, this doesn't really help us as we need to be able to trust > our toolchain and we shouldn't need to remerge stuff just to make sure. > I'll raise this to docs team and see what they say. > > /Fatih > > From: <opnfv-tech-discuss-boun...@lists.opnfv.org> on behalf of "Kunzmann, > Gerald" <kunzm...@docomolab-euro.com> > Date: Wednesday, 25 January 2017 at 11:28 > To: "opnfv-tech-discuss@lists.opnfv.org" > <opnfv-tech-discuss@lists.opnfv.org>, "opnfv-helpd...@rt.linuxfoundation.org" > <opnfv-helpd...@rt.linuxfoundation.org> > Subject: [opnfv-tech-discuss] [releng][helpdesk] Problem with > Git/Gerrit/Jenkins > > Dear all, > > Not sure how is the best team to address this to. We have noticed the > following issue: > > In Doctor project there had been two patches merged on the alarm comparison > table in Section 5.5.3: > > - https://gerrit.opnfv.org/gerrit/26715 > > - https://gerrit.opnfv.org/gerrit/27237 > > Few days later, we merged another patch addressing some editorial changes in > the document but NOT touching this alarm comparison table: > > - https://gerrit.opnfv.org/gerrit/26731 > > Unfortunately, this patch has overwritten the two patches merged earlier and > the alarm comparison table is back to the old version. > I would have expected > > a) that the patch is only changing the editorial fixes, and/or > > b) result in a “merge conflict” as parts of the repo had been changed > with the other patch. > > I had tried to revert the last patch 26731 in patch > https://gerrit.opnfv.org/gerrit/#/c/27531/ , but the document links provided > by jenkins-ci still show the old version of the table. > I would have expected that with the revert of the last patch, we would be > back to the version we had after having submitted patch 27237. > > Now, whilst I could manually fix this issue by providing another patch to > update the table, I am very concerned that this behavior may also cause > similar issues in other patch submissions that go unnoticed…. > > Can someone please have a look at this? > > Best regards, > Gerald > > =============================================== > Dr. Gerald Kunzmann, Manager > DOCOMO Communications Laboratories Europe GmbH Landsberger Strasse 312, 80687 > Munich > Tel: +49-89-56824-239 / Fax: +49-89-56824-300 > Web: http://www.docomoeurolabs.de<http://www.docomoeurolabs.de/> > > Geschäftsführer: Atsushi Takeshita, Dr. Thomas Walter, Hisahiro Hamahata, > Hiroyuki Oto, Kei Tonokura, Amtsgericht München, HRB 132976 > > > > _______________________________________________ > opnfv-tech-discuss mailing list > opnfv-tech-discuss@lists.opnfv.org > https://lists.opnfv.org/mailman/listinfo/opnfv-tech-discuss _______________________________________________ opnfv-tech-discuss mailing list opnfv-tech-discuss@lists.opnfv.org https://lists.opnfv.org/mailman/listinfo/opnfv-tech-discuss
