> On Jan 17, 2018, at 7:52 AM, Faseela K <[email protected]> wrote: > > Yes Sam. > We are not going to merge anything till the builds pass, and tests pass. > Luis’ mail had pointers that we will be able to soon build distribution, as > we are planning to remove projects pending the bump patches.
I did already that in int/distribution so that we can at least get something tested in CSIT. I think logical next step is: - For all projects that have not successfully merged the bump patch yet [1], work on that and report any problem. - For all projects that have already merged the bump patch and are part of the distribution (the remaining projects), check CSIT results of next distribution test [2] and report any issue you notice. For the second expect the following known issues: - Topology model version problem: https://lists.opendaylight.org/pipermail/openflowplugin-dev/2018-January/007912.html - Change of karaf prompt can make test very slow: https://lists.opendaylight.org/pipermail/release/2018-January/013412.html - Error booting some feature: it could be because of this: https://git.opendaylight.org/gerrit/#/c/67229/ [1] alto, eman, faas, gbp, memo, nic, of-config, p4plugin, packetcable, snmp4sdn, tsdr, unimgr, usc, vtn [2] https://jenkins.opendaylight.org/releng/view/distribution/job/integration-distribution-test-oxygen/115 BR/Luis > > Thanks, > Faseela > > From: Sam Hague [mailto:[email protected]] > Sent: Wednesday, January 17, 2018 6:54 PM > To: Faseela K <[email protected]> > Cc: Luis Gomez <[email protected]>; Abhijit Kumbhare <[email protected]>; > <[email protected]> <[email protected]>; Brady > Johnson <[email protected]>; Release ([email protected]) > <[email protected]>; Dayavanti Gopal Kamath > <[email protected]>; TSC <[email protected]>; > Stephen Kitt <[email protected]>; N Vivekanandan <[email protected]> > Subject: Re: [release] [OpenDaylight Discuss] [OpenDaylight TSC] odlparent 3 > / yangtools 2 migration > > Please hold off on merging non bump patches until the issues are fixed. > > Would like to reiterate that csit is failing horribly in many cases so > projects should hold off on merging any non bump patches. We don't want to > introduce further instability. This could all still be reverted if the > underlying issues can't be fixed as we are past our deadline for the > activity. > > On Jan 17, 2018 5:43 AM, "Faseela K" <[email protected]> wrote: > Thanks Luiz! > This should help us to get our pending patches in. > > Thanks, > Faseela > > From: Luis Gomez [mailto:[email protected]] > Sent: Wednesday, January 17, 2018 3:46 PM > To: Abhijit Kumbhare <[email protected]> > Cc: Brady Johnson <[email protected]>; Kitt, Stephen <[email protected]>; > tsc <[email protected]>; OpenDaylight Discuss > <[email protected]>; Release ([email protected]) > <[email protected]>; Faseela K <[email protected]> > Subject: Re: [OpenDaylight Discuss] [release] [OpenDaylight TSC] odlparent 3 > / yangtools 2 migration > > FYI, we are now building distribution with all projects except those pending > the bump patch: alto, eman, faas, gbp, memo, nic, of-config, p4plugin, > packetcable, snmp4sdn, tsdr, unimgr, usc and vtn: > > https://jenkins.opendaylight.org/releng/view/distribution/job/distribution-merge-oxygen/1569/ > > - Distribution-check should work after > https://git.opendaylight.org/gerrit/#/c/67229/ > - Autorelease-validate should work after removing above projects from > autorelease. > > BR/Luis > > > On Jan 16, 2018, at 11:26 PM, Abhijit Kumbhare <[email protected]> wrote: > > Thanks for all the work on this folks (too many to name: Stephen, Michael, > Luis, Jamo, Tom, Robert, Sam, Brady, Faseela, etc.). About the original > question Stephen from your email: > > "So the question now is what do we do. Following the agreed plan, we > should start reverting the patches (hopefully without messing up the > follow-up fixes). We could formulate alternative plans, but I’m not > sure what the community feels about that!" > > Do you guys (particularly Stephen & Michael since they may have been involved > on this the most) feel that we should decide this over the email thread or > does it make sense to wait till the TSC meeting? Of course a few folks like > Michael & Faseela may not be able to attend the TSC meeting - and can > continue to provide their response over this thread. > > On Tue, Jan 16, 2018 at 8:44 PM, Brady Johnson <[email protected]> wrote: > The SFC patch has merged successfully, so now the Netvirt version bump patch > can be merged. > > Regards, > > Brady > > > On Jan 16, 2018 21:32, "Luis Gomez" <[email protected]> wrote: > > On Jan 16, 2018, at 4:35 PM, Faseela K <[email protected]> wrote: > > Hi, > If there was an option to do such major impact changes on a different > branch, and then get it merged to main branch when things are stable > enough.(given that such a version bump-up is happening almost at the same > time as we are going to do a branch cut), would have been far better. > > I am not sure how a development branch would have helped here. In general to > catch issues related to upstream bump before projects perform the bump we can: > > - Prepare patches for everybody and use recheck with patch dependency: this > worked pretty well for me and helped identifying some issues. Problem is > someone has to prepare the patches and this method does not generate a > distribution to test. > - Prepare patches for everybody and use multi-patch job: It is similar to the > above but builds in a different way, because of that it cannot be used for a > major bump change as it is today but this is fixable and this job does > actually generate a distribution and trigger CSIT if required. > - Have an automatic upstream bump job > (https://jira.opendaylight.org/projects/RELENG/issues/RELENG-69): This is not > there yet but I believe it is very feasible once we have an streamlined > distribution with stable autorelease build. > > Now I have to say that nothing of the above really works if all projects > involved in the change are not actively looking at the issues exposed by the > above methods, fixing them and reporting them to upstream when they cannot. I > think this last is what is really failing today and explains why upstream and > ultimately TSC decided to break everybody for a while rather than going > through the painful (and slow) process of asking one by one all projects to > check the bump change. > > BR/Luis > > > But would like to tell that some of the issues we had with the previous > versions seem to go off once we got a stable genius distribution(eg : ssh > related problems as sshd got upgraded). > Thanks, > Faseela > > From: [email protected] > [mailto:[email protected]] On Behalf Of Luis Gomez > Sent: Wednesday, January 17, 2018 5:58 AM > To: Jamo Luhrsen <[email protected]> > Cc: Stephen Kitt <[email protected]>; OpenDaylight Discuss > <[email protected]>; Release ([email protected]) > <[email protected]>; tsc <[email protected]> > Subject: Re: [release] [OpenDaylight TSC] odlparent 3 / yangtools 2 migration > > FYI I am trying this patch in sandbox for the issue you observed in CSIT: > > https://git.opendaylight.org/gerrit/#/c/67224/ > > BR/Luis > > > On Jan 16, 2018, at 1:56 PM, Jamo Luhrsen <[email protected]> wrote: > > good idea Sam. I'll try that. > > btw, you could do a curl on restconf/modules and get a 200 right? > > JamO > > On 01/16/2018 01:55 PM, Sam Hague wrote: > > I loaded that distro locally and did feature:install odl-netvirt-openstack. > That worked fine. Maybe we can use the patch you > added recently to not use the bootfeatures? > > On Jan 16, 2018 4:02 PM, "Jamo Luhrsen" > <[email protected]<mailto:[email protected]>> wrote: > > Thanks Luis, > > that fixed it, and I ran new jobs already. ODL did not come "UP", so we > have something to fix still. > > I'll be debugging, but something weird ended up in the features cfg file. > there is an UUID in the featuresBoot and featuresRepositories lines > > > https://jenkins.opendaylight.org/releng/user/jluhrsen/my-views/view/netvirt%20csit/job/netvirt-csit-1node-openstack-ocata-gate-stateful-oxygen/638/ > > <https://jenkins.opendaylight.org/releng/user/jluhrsen/my-views/view/netvirt%20csit/job/netvirt-csit-1node-openstack-ocata-gate-stateful-oxygen/638/> > > JamO > > > > On 01/16/2018 12:38 PM, Luis Gomez wrote: > > I mean the format of the URL so that script can take it :) > > Just add 2 folders (/karaf/0.8.0-SNAPSHOT) in you HTTP server and it should > work, for example: > > http://www.luhrsenlounge.net/odl/karaf/0.8.0-SNAPSHOT/karaf-0.8.0-SNAPSHOT.zip > > <http://www.luhrsenlounge.net/odl/karaf/0.8.0-SNAPSHOT/karaf-0.8.0-SNAPSHOT.zip> > > > > On Jan 16, 2018, at 12:35 PM, Jamo Luhrsen <[email protected] > <mailto:[email protected]>> wrote: > > > > On 01/16/2018 12:27 PM, Luis Gomez wrote: > > I think you have to use something like this in the > BUNDLE_URL: > http://www.luhrsenlounge.net/odl/*karaf/0.8.0-SNAPSHOT*/karaf-0.8.0-SNAPSHOT.zip > > <http://www.luhrsenlounge.net/odl/*karaf/0.8.0-SNAPSHOT*/karaf-0.8.0-SNAPSHOT.zip> > > > hmmm... the BUNDLE_URL is right, but our deployment script is having trouble > using > it. I'm looking. > > On Jan 16, 2018, at 11:51 AM, Jamo Luhrsen > <[email protected]<mailto:[email protected]> > <mailto:[email protected] <mailto:[email protected]>>> wrote: > > > Sam built a local distro with the remaining unmerged patches > and I am trying some netvirt CSIT with that: > > > https://jenkins.opendaylight.org/releng/user/jluhrsen/my-views/view/netvirt%20csit/job/netvirt-csit-1node-openstack-ocata-gate-stateful-oxygen/637 > > <https://jenkins.opendaylight.org/releng/user/jluhrsen/my-views/view/netvirt%20csit/job/netvirt-csit-1node-openstack-ocata-gate-stateful-oxygen/637> > > > > > https://jenkins.opendaylight.org/releng/user/jluhrsen/my-views/view/netvirt%20csit/job/netvirt-csit-1node-openstack-ocata-gate-stateful-snat-conntrack-oxygen/38 > > <https://jenkins.opendaylight.org/releng/user/jluhrsen/my-views/view/netvirt%20csit/job/netvirt-csit-1node-openstack-ocata-gate-stateful-snat-conntrack-oxygen/38> > > > > JamO > > > > On 01/16/2018 11:07 AM, Michael Vorburger wrote: > > On Tue, Jan 16, 2018 at 7:29 PM, Brady Johnson > <[email protected]<mailto:[email protected]> > <mailto:[email protected] <mailto:[email protected]>>> wrote: > > > We have indeed made good progress and it looks like we're close. Should > we at least give it until the end of the > day today? > > > Related to this, since the merges started I havent been able to build SFC > since Friday when I sent this email > [0] and I > > still cant today. So this is something we need to resolve soon to avoid > schedule slips. > > [0] > https://lists.opendaylight.org/pipermail/release/2018-January/013337.html > <https://lists.opendaylight.org/pipermail/release/2018-January/013337.html> > > <https://lists.opendaylight.org/pipermail/release/2018-January/013337.html > > <https://lists.opendaylight.org/pipermail/release/2018-January/013337.html>> > > > > Regards, > > *Brady Johnson* > [email protected]<mailto:[email protected]> > <mailto:[email protected]<mailto:[email protected]>> > > > <https://twitter.com/inocybetech<https://twitter.com/inocybetech>><http://www.inocybe.com/ > > <http://www.inocybe.com/>><https://www.linkedin.com/company/2661537?trk=tyah&trkInfo=clickedVertical%3Acompany%2CclickedEntityId%3A2661537%2Cidx%3A1-1-1%2CtarId%3A1441300264767%2Ctas%3Ainocybe > > <https://www.linkedin.com/company/2661537?trk=tyah&trkInfo=clickedVertical%3Acompany%2CclickedEntityId%3A2661537%2Cidx%3A1-1-1%2CtarId%3A1441300264767%2Ctas%3Ainocybe>>Screenshot > > 2017-02-14 at 10.43.55 AM.png > <https://www.youtube.com/channel/UC9uUWABdPR0Je9Du_15FCkw > <https://www.youtube.com/channel/UC9uUWABdPR0Je9Du_15FCkw>> > > > > > On Tue, Jan 16, 2018 at 8:31 AM, Stephen Kitt > <[email protected]<mailto:[email protected]> > <mailto:[email protected] <mailto:[email protected]>>> wrote: > > > Hi, > > In last week’s TSC call we agreed to move forward with merging the > odlparent 3 / yangtools 2 migration patches, with a timebox — we were > supposed to be able to run NetVirt CSIT yesterday. Obviously that > deadline has gone whizzing past and we’re still some way away from > running NetVirt CSIT; I’m not sure how far away... Strictly speaking, > all the NetVirt dependencies now pass verify (thanks to lots of work > by > Tom, Robert and Michael on Neutron in particular), so we could give it > a shot once the patches are merged, but I also knows that OVSDB fails > its IT so I expect NetVirt CSIT won’t pass. > > So the question now is what do we do. Following the agreed plan, we > should start reverting the patches (hopefully without messing up the > follow-up fixes). We could formulate alternative plans, but I’m not > sure what the community feels about that! > > Anyway, I’m not trying to suggest a concrete solution with this email, > I want to start the discussion. > > > my personal view is that the sane thing to do would be to revert and finish > this up on a branch in parallel, to unblock > folks, because the timeline proposed in > https://lists.opendaylight.org/pipermail/release/2018-January/013336.html > > <https://lists.opendaylight.org/pipermail/release/2018-January/013336.html> > was missed, > > and IMHO it's likely at least several more days until things fully stabilize. > Through all the various 3rd party > upgrades and > > yangtools changes etc. that went into this big bump, as of right there still > has never been a full netvirt CSIT > successfully > > run proving non-regression, yet; that's going to happen in the coming days. > > But I believe this view is not widely shared... ;-) Part of the reasons for > that seem to include issues such as > that the > > multipatch job apparently doesn't work, so it is hard to test change accross > projects easily. I'm not sure that > "let's just > > merge and figure it out as we go along" is the best answer to that problem.. > but if others don't have a problem > with the > > "broken world" for a number of days, then so be it! > > > Regards, > > -- > Stephen Kitt > Principal Software Engineer, Office of the CTO > Red Hat > > _______________________________________________ > release mailing list > [email protected]<mailto:[email protected]> > <mailto:[email protected] > <mailto:[email protected]>> > > https://lists.opendaylight.org/mailman/listinfo/release > <https://lists.opendaylight.org/mailman/listinfo/release> > <https://lists.opendaylight.org/mailman/listinfo/release > <https://lists.opendaylight.org/mailman/listinfo/release>> > > > > > _______________________________________________ > TSC mailing list > [email protected]<mailto:[email protected]> > <mailto:[email protected] > <mailto:[email protected]>> > > > https://lists.opendaylight.org/mailman/listinfo/tsc<https://lists.opendaylight.org/mailman/listinfo/tsc> > > <https://lists.opendaylight.org/mailman/listinfo/tsc<https://lists.opendaylight.org/mailman/listinfo/tsc>> > > > > > > _______________________________________________ > release mailing list > [email protected]<mailto:[email protected]> > https://lists.opendaylight.org/mailman/listinfo/release<https://lists.opendaylight.org/mailman/listinfo/release> > > _______________________________________________ > TSC mailing list > [email protected]<mailto:[email protected]> > https://lists.opendaylight.org/mailman/listinfo/tsc<https://lists.opendaylight.org/mailman/listinfo/tsc> > > > _______________________________________________ > TSC mailing list > [email protected] <mailto:[email protected]> > > https://lists.opendaylight.org/mailman/listinfo/tsc<https://lists.opendaylight.org/mailman/listinfo/tsc> > > > > _______________________________________________ > release mailing list > [email protected] > https://lists.opendaylight.org/mailman/listinfo/release > > > > _______________________________________________ > Discuss mailing list > [email protected] > https://lists.opendaylight.org/mailman/listinfo/discuss > > > > > _______________________________________________ > release mailing list > [email protected] > https://lists.opendaylight.org/mailman/listinfo/release > _______________________________________________ Discuss mailing list [email protected] https://lists.opendaylight.org/mailman/listinfo/discuss
