13:59:52 <anipbu> #startmeeting release 13:59:52 <odl_meetbot> Meeting started Wed Jul 13 13:59:52 2016 UTC. The chair is anipbu. Information about MeetBot at http://ci.openstack.org/meetbot.html. 13:59:52 <odl_meetbot> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 13:59:52 <odl_meetbot> The meeting name has been set to 'release' 14:00:12 <anipbu> #topic Upgrade IETF Types 14:00:45 <zxiiro> Hello, Thanh is here 14:01:16 <anipbu> #info We will be merging patches for the Upgrade ietf-{inet,yang}-types to 2013-07-15 14:01:32 <anipbu> #link https://wiki.opendaylight.org/view/Weather#Upgrade_ietf-.7Binet.2Cyang.7D-types_to_2013-07-15 <--- Weather Item for upgrading IETF Types 14:01:47 <zxiiro> So what's the plan? do I need to merge everything at once? or do we do it one at a time in some order? 14:02:00 <zxiiro> lori: ^ 14:02:00 <anipbu> #info Currently we have +2/+1 from all projects except SNMP 14:02:18 <anipbu> #link https://docs.google.com/spreadsheets/d/1zImtd764e-hOgJAxoJKl85fxHCPu2agLfqsBtf13zQY/edit#gid=778533050 <--- Tracking approvals from projects. 14:02:27 <lori> zxiiro: we do it in order IMO 14:02:42 <lori> we can merge mdsal, controller, and openflowjava at once 14:02:48 <lori> since they are independent 14:02:56 <anipbu> #info we will be merging patches based on the order in column "G" 14:03:05 <lori> once their merge jobs are successful 14:03:13 <lori> we can merge openflowplugin 14:03:17 <lori> and so on 14:03:18 <zxiiro> ok, i think snmp is a pretty close to leaf project right? 14:03:24 <anipbu> lori: zxiiro: Good Morning! 14:03:27 <zxiiro> so it might be ok for us to start without them? 14:03:37 <lori> anipbu: zxiiro: good morning :) 14:03:45 <zxiiro> good mornight 14:03:46 <anipbu> yes, SNMP can be ignored (not merged) without impact to other projects 14:03:47 <lori> zxiiro: IMO yes 14:04:14 * colindixon watches the magic happen 14:04:32 <lori> I will be building stuff locally in parallel, in case any of the patches cause issues I can submit a fix 14:04:32 <zxiiro> ok so patch 1 is mdsal, it failed verified. 14:04:36 <zxiiro> is that expected? 14:04:45 <lori> it failed distro-check I think 14:04:57 <lori> beucase distro-check detected it would break distribution 14:05:04 <zxiiro> ah ok yeah that would be expected then 14:05:09 <zxiiro> cool so we can ignore the vote 14:05:10 <lori> so yes, it should be ok to merge 14:05:26 <lori> same for controller 14:05:30 <zxiiro> alright doing it soon... just gotta enable permissions for myself 14:05:36 <lori> ok 14:06:07 <anipbu> #info status: Thanh is merging MDSAL first right now 14:07:28 <zxiiro> ah we're getting a play by play ;) 14:08:00 <zxiiro> ok mdsal merged. Do we need to wait for the merge job? 14:08:07 <anipbu> it's like EPSN :) 14:08:23 <anipbu> except, do they watch ESPN in Canada? 14:08:35 <zxiiro> yes 14:08:41 <zxiiro> depends on the sport though 14:08:42 <lori> no, please merge controller and openflowjava as well 14:08:57 <lori> and then we wait for all merge jobs to finish 14:09:27 <zxiiro> lori: ok controller, and openflowjava all merged 14:09:37 <lori> awesome, thanks 14:09:56 <anipbu> #info status: merging CONTROLLER right now 14:10:07 <anipbu> #info status: merging OPENFLOWJAVA right now 14:10:18 <colindixon> lori, anipbu: if we wait for merge jobs to finish, we'll be here all day 14:10:59 <lori> colindixon: well, if we want correct artifacts to be published, and minimize breakage, I don't see a better way 14:11:13 <anipbu> colindixon: what is your recommendation? 14:11:16 <lori> for example, if we merge openflowplugin now, its merge job will fail 14:11:35 <lori> and then until all projects aresorted out, the confusion will probably be bigger 14:11:45 <lori> then just merging in order 14:12:10 <vrpolak> Would openflowjava build need artifactId ietf-inet-types-2013-07-15 (from mdsal build) to be present in Nexus to succeed? 14:12:25 <lori> zxiiro: is there a way to prioritize the merge jobs we trigger now on Jenkins 14:12:39 <lori> vrpolak: openflowjava no, openflowplugin yes 14:12:55 <zxiiro> lori: the priority sorter plugin already priorities merge jobs and the queue is short so they are all jumping to the beginning of the queue 14:13:08 <lori> zxiiro: excellent 14:13:55 <colindixon> so, that's fine, I guess the tree is wide enough that it won't take too long 14:14:04 <colindixon> do we know the depth of the patch tree? 14:14:11 <vrpolak> lori: What about this then? https://git.opendaylight.org/gerrit/#/c/40787/2/openflow-protocol-api/pom.xml 14:15:10 <lori> vrpolak: those artifacts were available for a long time 14:15:24 <lori> the mdsal patch only changes the network-topology patches 14:15:33 <lori> which openflowplugin uses 14:16:15 <lori> colindixon: I don't know exactly the depth of the tree, but as you say, it shouldn't too deep 14:16:36 <vrpolak> lori: Oh, ok. I forgot about them being already there. 14:21:11 <zxiiro> fyi i'm gonna cancel the integration jobs to keep teh queue low and since we're watching and merging patches to projects anyway we'll notice if something breaks without that job 14:21:24 <colindixon> zxiiro: cool, thanks 14:22:48 <zxiiro> i think openflowjava should merge fairly quickly but controller and mdsal take about 40 minutes if i recall 14:23:45 <zxiiro> yeah 37 minutes for mdsal and 53 minutes for controller 14:29:58 <vrpolak> Oh, so this simultaneous upgrade is done because we do not want to create artificial new revision of network-topology and similar, but we still want the models with updated types. 14:34:06 <colindixon> {mdsal, controller} => {aaa, bgpcep, netfconf, neutron, openflowjava, sxp, usc} => {of-config, openflowplugin} => {didm, l2switch, lacp, nemo, ovsdb, sdninterfaceapp, topoprocessing} => {sfc, genius, unimgr, vtn} => {netvirt, groupbasedpolicy} => nic => vpnservice; snmp hasn't answered and might block tsdr? 14:34:15 <vrpolak> (And we do not want different revisions to be present at once in ODL anyway, because restconf cannot distinguish between them.) 14:34:16 <colindixon> that's my topological sort 14:34:38 <colindixon> depth 8 14:34:55 <colindixon> I think that's right, but could use double checking 14:38:29 <colindixon> anipbu: if you look at the above sort, I think we can do the things in {}s in parallel which will fix the most painful waiting 14:42:41 <lori> colindixon: yes, I also think we could do things in parallel 14:42:50 <lori> I just didn't do the above exceprcise :) 14:43:34 <anipbu> colindixon: okay, I put your parallel ordering in row 75 for us to follow and track (https://docs.google.com/spreadsheets/d/1zImtd764e-hOgJAxoJKl85fxHCPu2agLfqsBtf13zQY/edit#gid=778533050) 14:44:34 <anipbu> lori: based on colindixon's analysis, once {mdsal,controler} is complete, do you think we should merge {aaa, bgpcep, netfconf, neutron, openflowjava, sxp, usc} altogether? 14:44:47 <lori> aaa and openflowjava are merged 14:45:00 <lori> and yes, we can merge the others at the same time 14:45:19 <colindixon> I'm just trying to make sure lori sleeps tonight :p 14:45:27 <lori> :) thanks! 14:46:05 <colindixon> and it's possible the dependencies there are not at the model-level, but it's probably good to be safe 14:46:06 <lori> mdsal finished merge 14:46:55 <colindixon> we're waiting on this: https://jenkins.opendaylight.org/releng/job/controller-merge-boron/414/ 14:48:46 <colindixon> currently at 36 of ~53 minutes 14:48:56 <colindixon> ETA: UTC 15:05 14:49:27 <colindixon> it's archiving now 14:49:29 <colindixon> so maybe sooner 14:49:58 * colindixon watches the paint dry 14:50:40 <colindixon> uploading to nexus now 14:51:24 <lori> :) 14:51:39 <anipbu> colindixon: a watched pot never boils 14:51:49 <anipbu> :) 14:52:48 <lori> not good 14:52:50 <lori> merge failed 14:52:52 <colindixon> ugh 14:53:05 <colindixon> org.apache.maven.artifact.deployer.ArtifactDeploymentException: Failed to deploy artifacts: Could not transfer artifact org.opendaylight.controller:config-manager-facade-xml:jar:javadoc:0.5.0-20160713.145009-352 from/to opendaylight-snapshot (https://nexus.opendaylight.org/content/repositories/opendaylight.snapshot/): Connection timed out 14:53:23 <lori> wow 14:53:30 <lori> such a trivial thing :( 14:53:42 <colindixon> I just kicked a remerge 14:53:52 <colindixon> which is going to take a while, unless zxiiro has some magic 14:54:01 <lori> 40 min. wasted 14:54:26 <colindixon> and the remerge isn't kicking 14:54:59 <anipbu> Could not transfer artifact; Connection timed out.... is this related to the artifact size discussion on the mailing list? 14:55:21 <lori> no, the error was connection timed out 14:55:23 <colindixon> anipbu: I don't think so, I think it's just that was a big artifact and nexus has been flaky 14:55:33 <colindixon> https://jenkins.opendaylight.org/releng/view/controller/job/controller-merge-boron/ 14:55:38 <colindixon> it's the remerge is queued 14:59:39 <colindixon> https://jenkins.opendaylight.org/releng/job/controller-merge-boron/415/ 15:00:03 <colindixon> unless zxiiro has other ideas? should we kick off 2-3 copies of it just to make sure one finishes? 15:01:40 <zxiiro> let me look 15:02:46 <zxiiro> doh connection timed out :( 15:02:48 <zxiiro> tykeal: ^ 15:02:59 <zxiiro> hopefully the next one works 15:03:16 <colindixon> zxiiro: could we run two merge jobs at the same time 15:03:19 <colindixon> so that at least one works? 15:03:33 <zxiiro> colindixon: I could modify the job to allow more than one merge 15:05:16 <zxiiro> there a 2nd one should be running once it acquires a slave 15:05:43 <colindixon> zxiiro: for this process it seems like a good idea, just today 15:07:27 <zxiiro> openflowjava and mdsal were success 15:07:43 <zxiiro> any other patches we can merge considering? or do we need to wait for controller? 15:08:35 <lori> zxiiro: from what I know we need to wait for controller 15:08:51 <lori> although that may not be accurate 15:09:00 <colindixon> I think everything else depends on controller 15:09:01 <colindixon> let me check 15:09:22 <lori> however, I think the changes in neutron and netconf are NOT dependent on the changes in controller 15:09:28 <lori> from this particular patch set 15:09:33 <lori> so we could merge them 15:09:40 <lori> and if they fail, we can just remerge 15:09:46 <colindixon> fair 15:09:57 <colindixon> next doesn't depend on controller, but also doesn't matter 15:10:44 <zxiiro> I think we really need to stop this process of allowing every project to trigger jobs off of offset0 projects. I've noticed every time our queue gets destroyed it's because of an offset 0 project merging a patch. I don't think we need all 80 ODL projects to verify that a controller branch broke the world. 15:11:20 <colindixon> zxiiro: we should talk about that 15:13:04 <lori> let's merge netconf, neutron, they don't use the network-topology model that the controller patch changes 15:13:14 <zxiiro> lori: will do 15:13:24 <lori> thanks! 15:15:36 <anipbu> #info status: merging NETCONF right now 15:15:43 <anipbu> #info status: merging NEUTRON right now 15:17:42 <zxiiro> anipbu: should livestream on twitch, make some of that ad revenue :D 15:17:47 <lori> :) 15:17:59 <anipbu> :) 15:19:38 <zxiiro> lori: do we need unit testing? just thought that if we ran the jobs with -DskipTests on purpose we could speed this whole thing up 15:20:08 <lori> zxiiro: I would be more confident that we don't break anything if we keep tests 15:20:19 <zxiiro> ok just a thought 15:20:50 <lori> quite a few projects raised concerns that they didn't see a passing verify job 15:21:36 <lori> let's merge usc too 15:21:44 <lori> it doesn't depend on controller 15:21:59 <lori> in fact, it has a passing verfiy 15:22:11 <lori> same with sxp 15:22:35 <colindixon> lori: thanks for checking that 15:22:49 <colindixon> do we have other things that can pass? 15:22:52 <lori> yeah, I want to make sure I sleep too :) 15:22:58 <lori> no, just these two 15:23:18 <lori> but I will be going through patches to see patch-level dependencies 15:23:23 <lori> not project-level 15:23:27 <lori> see where we can optimize 15:26:13 <anipbu> #info status: merging USC right now 15:26:40 <anipbu> zxiiro: we should merge SXP 15:28:46 <zxiiro> ok doing the 2 now 15:28:54 <vrpolak> lori: "network-topology model that the controller patch changes" Netconf uses network-topology. But that was not changed by controller, only network-inventory was changed. Which in turn is not used by more projects, such as Bgpcep (afaik). 15:29:22 <zxiiro> ah An already did usc 15:29:28 <anipbu> zxiiro: i merged usc 15:29:30 <zxiiro> SXP is now merged 15:29:51 <anipbu> #info status: merging SXP right now 15:30:00 <lori> vrpolak: you're right network-inventory 15:31:15 <lori> but I see classes with "flowspec" in it in the bgpcep patch 15:31:24 <lori> taking a loot to see if it's ofp related 15:32:23 <lori> I'll build bgpcep locally and report back 15:32:44 <lori> we may be able to merge as vrpolak suggests 15:33:48 <vrpolak> lori: Module flowspec does not depend on anything related to openflow (or inventory). 15:34:05 <lori> excellent, let's merge then 15:34:13 <lori> it takes a lot to build, so we'll save time 15:34:26 <anipbu> lori: vrpolak: does this mean we can merge bgpcep? 15:35:09 <milfabia> yes we can, bgpcep was depending on md-sal patch 15:35:24 <vrpolak> anipbu: I believe yes, we can merge Bgpcep now. 15:35:29 <anipbu> zxiiro: please merge bgpcep 15:36:38 <colindixon> I'm going to go get lunch, please text/call me (301.455.5083) if you need anything 15:36:54 <zxiiro> anipbu: merged 15:37:05 <anipbu> controller built successfully: https://jenkins.opendaylight.org/releng/job/controller-merge-boron/415/ 15:37:15 <lori> neutron too 15:37:24 <lori> we can merge openflowplugin 15:37:25 <anipbu> zxiiro: should we kill job 416? 15:37:30 <lori> the major dependency for everyone else 15:37:50 <anipbu> #info status: merging BGPCEP right now 15:38:42 <anipbu> #info Phase One Complete: merge job complete and artifacts uploaded to nexus for MDSAL, CONTROLLER 15:38:46 <zxiiro> ok so merging openflowplugin, and neutron? 15:39:13 <anipbu> neutron should already have been merged. 15:39:28 <anipbu> yes, please merge openflowplugin 15:39:42 <lori> openflowpplugin, yes 15:39:52 <lori> neutron finished merge successfully 15:40:00 <zxiiro> ok done 15:41:03 <anipbu> #info status: merging OPENFLOWPLUGIN right now 15:43:19 <lori> netconf finished successfully 15:43:44 <anipbu> sxp finished successfully as well 15:45:05 <anipbu> usc completed successfully as well 15:48:23 <anipbu> lori: what are your thoughts on of-config? does it have any dependency on bgpcep? otherwise, could we move it along? 15:48:44 <lori> of-config seems to have no +2 15:49:06 <lori> well, the original patch does 15:49:12 <lori> the rebased pacthes not 15:49:21 <lori> is it ok to merge anyway? 15:50:13 <lori> AFAIK, no-one depends on of-config, so we can wait for a committer to merge it 15:50:24 <lori> maybe we can just trigger a reverify? 15:50:30 <lori> which should pass now 15:50:32 <anipbu> lori: yes. if they +2 original patch, we can merge 15:50:36 <lori> ok 15:50:53 <anipbu> as long as you are okay with the rebased patch to merge, we should go ahead and merge. 15:51:34 <lori> yes, I'm ok with that, let's merge 15:51:49 <anipbu> zxiiro: please merge of-config 15:55:15 <lori> it looks like netvirt merged some intrusive patch again, will have to rebase it, since gerrit reports merge conflict 15:55:47 <anipbu> lori: okay 15:56:32 <anipbu> lori: MISSING: I noticed that faas and alto is missing in our merge ordering (row 75). In which phase do you think we should merge these two projects? 15:56:58 <lori> we'll do in the order shown by colindixon's analysis 15:59:05 <anipbu> lori: colindixon analysis does not include faas and alto 15:59:48 <lori> ah 15:59:51 <lori> let me check 16:00:59 <zxiiro> anipbu: sorry missed your message doing it now 16:01:13 <anipbu> zxirro: thanks! 16:01:39 <lori> we can include both faas and alto in the next group, after openflowplugin 16:03:31 <anipbu> lori: thanks! 16:06:31 <anipbu> lori: zxiiro: Since openflowplugin is a bottleneck for the next set of patches (phase 3), a possible failure would delay us significantly. How do you guys feel about running a second backup merge job for openflowplugin, just in case it fails? 16:07:00 <lori> yeah, that sounds good 16:07:08 <lori> also, it takes quite a lot 16:07:24 <lori> and then, we should do the same for ovsdb, which will be the bottleneck for the next group 16:10:05 <zxiiro> yeah once you reach openflowplugin it seems like everything is bottlenecked on the project before it 16:10:27 <zxiiro> i think it's like openflowplugin -> sfc -> ovsdb -> vpnservice -> vtn or somethign 16:10:50 <zxiiro> I don't remember exactly but I have this problem everytime we version bump waiting for the last fewt hat can't be done in parallel 16:11:17 <zxiiro> ok i launched a 2nd openflowplugin job 16:11:27 <lori> zxiiro: thanks! 16:11:29 <anipbu> zxiiro: thanks! 16:15:17 <lori> of-config successful' 16:16:49 <zxiiro> i wish tests ran faster heh 16:16:58 <zxiiro> or at least printed soemthing 16:17:06 <anipbu> of the projects so far, it seems controller, bgpcep, and openflowplugin have taken a longer time compared to other projects. 16:17:17 <zxiiro> the long wait after it prints "SingleFeatureTest" makes me nervous 16:17:41 <zxiiro> yes they are the biggest 16:17:45 <zxiiro> also mdsal 16:17:58 <zxiiro> and ovsdb and netvirt too if i recall 16:21:16 <colindixon> zxiiro: it did that both times for the controller 16:21:19 <colindixon> how goes its/ 16:21:20 <colindixon> ? 16:21:59 <anipbu> are the integration jobs necessary for today? would we free up more resources if we kill (just for today) the bgpcep-integration-boron? 16:22:01 <zxiiro> things are going well, i think we're waiting on openflowplugin at the moment 16:22:14 <zxiiro> anipbu: yeah i've been kililng integration jobs as they pop up 16:22:29 <colindixon> did we start a 2nd ofp merge job? 16:22:31 <anipbu> colindixon: we have completed phase one. Phase two is waiting on bgpcep. Phase three is waiting on openflowplugin. 16:22:39 <zxiiro> colindixon: yes 16:25:51 <anipbu> jamoluhrsen. could we also kill performance tests (just for today). I mean, is it absolutely necessary that we run openflowplugin-csit-1node-periodic-longevity-only-beryllium today for example? 16:26:10 <jamoluhrsen> kill it 16:26:12 <anipbu> jamoluhrsen: just trying to think of ways to free up more resources 16:26:31 <colindixon> are we resource constrained? 16:26:34 <jamoluhrsen> things are hurting in so many ways right now, I think it's ok to kill it. 16:26:37 <colindixon> or dependency constrained? 16:26:46 <jamoluhrsen> LuisGomez: ok if we kill OFP longevity/perf CSIT? 16:26:54 <lori> right now I'd say dependency 16:27:06 <lori> but once ofp is merged 16:27:15 <lori> I mean the merge job passes 16:27:20 <zxiiro> resources on infra is fine 16:27:21 <lori> we will have several projects 16:27:25 <LuisGomez> sure 16:27:28 <lori> that we can merge at once 16:27:36 <zxiiro> the robot queue does not affect the merge queue 16:27:41 <zxiiro> they are 2 separate buckets 16:27:49 <lori> ah, perfect 16:28:00 <zxiiro> the only things is kiling off periodic and integration jobs 16:28:04 <zxiiro> which i've been doing as they pop up 16:28:12 <anipbu> zxiiro: ah okay, so killing those performance tests won't have any added benefits. 16:28:26 <zxiiro> merge , verify, integration, periodic, all share teh same 30 max VMs 16:28:27 <jamoluhrsen> ofp longevity is killed. 16:28:47 <jamoluhrsen> netconf is running some scale CSIT if we need to find something else that's probably using many resources 16:28:58 <zxiiro> plus merge jobs and verify jbos jump to the front of the queue when they join the queue due to prioritization 16:29:03 <colindixon> do we know if tsdr depends on snmp for the things it uses types for? 16:29:44 <colindixon> so, who's catching Pokémon while waiting for merge jobs to finish? 16:29:56 <zxiiro> jamoluhrsen: but we should definitely talk about csit triggers at some point though. we still have that problem where every job is triggering off of offset0 16:30:05 <zxiiro> LuisGomez: ^ 16:30:38 <lori> colindixon: let me take a quick look at the patches 16:30:39 <anipbu> colindixon: someone used my username for Pokemon Go. I need to get a less common username. 16:30:41 <zxiiro> I think we need to come up with a different strategy, we can't have almost every ODL project trigger on every offset0 job merge 16:30:45 <LuisGomez> right, we can talk tomorow in our call 16:31:05 <colindixon> I think that given that we're not gating on the results, there's limited value to running integration jobs on every patch? 16:31:17 <colindixon> maybe just once a day (if one of your dependencies changed) 16:31:22 <colindixon> is the second part doable? 16:31:26 <jamoluhrsen> anipbu: that sounds fishy. I'm sure you are out there every day finding those virtual pokemons 16:31:27 <zxiiro> One idea is to maybe reverse that. Have the offset0 projects pick a few projects that we believe cover most of the ecosystem and trigger only those jobs 16:31:42 <jamoluhrsen> zxiiro, LuisGomez: csit triggers convo is on my list now 16:31:48 <zxiiro> ok cool 16:33:04 <zxiiro> colindixon: I don't have pokemon yet, they ahven't launched in canada officially (although sideloading is an option ) 16:34:42 <colindixon> zxiiro, anipbu: I figured I needed to sign up to figure out what the hubbub was about, it's more interesting than fun 16:34:48 <zxiiro> as someone who contributes to openstreetmap though, I wish Pokemon Go's data collection helped improve OSM rather than Google Maps 16:34:48 <colindixon> but it is interesting 16:35:52 <anipbu> lori: zxiiro: if openflowplugin completes BUT bgpcep does not complete, then I think we can go ahead and merge phase four except for sdninterfaceappp (which depends on bgpcep). 16:36:09 <lori> anipbu: yes 16:38:32 <lori> colindixon: I think tsdr doesn depend on snmp, since files from this bundle are changed: https://git.opendaylight.org/gerrit/gitweb?p=tsdr.git;a=blob;f=collectors/snmp-data-collector/pom.xml;h=2558db3c947080f40f53eae0720454ab2e28404c;hb=5eef95850e9ae50d4c346688071bd1b4123376e4 16:38:44 <lori> colindixon: and snmp is declared as dependency 16:41:11 <anipbu> colindixon: yes, tsdr depends on snmp for their snmp collector. I've marked their project as red (do not merge) 16:42:47 <anipbu> snmp was planning to drop from boron, but at the last minute Sharon Aicler from Cisco stepped up to take over the reigns for snmp. I have reach out to Sharon Aicler but have not heard back. 16:45:05 <colindixon> OK. 16:45:19 <colindixon> maybe edwarnicke could ping Sharon 16:49:01 <zxiiro> openflowplugin is done 16:49:04 <zxiiro> what next? 16:49:28 <lori> ovsdb 16:49:29 <zxiiro> bgpcep too 16:49:44 <lori> this group: {didm, l2switch, lacp, nemo, ovsdb, sdninterfaceapp, topoprocessing} 16:49:50 <lori> but start with ovsdb 16:49:55 <vrpolak> zxiiro: "pick a few projects" I would nominate Bgpcep, Openflowplugin, Lispflowmapping and Ovsdb (in decreasing order of my vague recollection of frequency of breakage detection). 16:49:57 <lori> and please 2 merge jobs for ovsdb 16:50:09 <zxiiro> ok ovsdb merged 16:50:36 <lori> thanks 16:51:12 <lori> next should be nemo, since there is some project depending on it 16:51:31 <lori> and then the rest in whatever order 16:51:38 <zxiiro> ok merge x2 started just waiting for an instance 16:51:42 <anipbu> zxiiro: please kill the extra openflowplugin job: https://jenkins.opendaylight.org/releng/job/openflowplugin-merge-boron/257/ 16:52:20 <zxiiro> done 16:53:09 <anipbu> #info status: merging OVSDB right now 16:53:32 <zxiiro> gonna step out to grab a lunch and come back. colindixon also has the powers to merge +2 patches 16:54:14 <anipbu> bgpcep has not finished yet: https://jenkins.opendaylight.org/releng/job/bgpcep-merge-boron/264/ 16:54:48 <lori> zxiiro: enjoy! 16:55:00 <anipbu> zxiiro: have fun with lunch! 16:55:00 <zxiiro> anipbu: it's done, it's just uploading it's maven site 16:55:06 <zxiiro> anipbu: which for our purposes you can ignore 16:55:21 <anipbu> zxiiro: okay. 16:55:51 <zxiiro> anipbu: oh wait no you're right, the maven-site isn't a publisher step :( 16:55:56 <zxiiro> anipbu: so the upload of artifacts didn't happen yet 16:56:05 <zxiiro> we'll have to move that in the future 16:56:19 <zxiiro> make it more efficient 16:56:22 <anipbu> lori: can we merge nemo now? I am comitter and can merge this patch. 16:56:25 <lori> yes 16:56:27 <colindixon> ok 16:56:28 <colindixon> I'm doing it 16:57:02 <colindixon> actually, I can't merge 16:57:03 <colindixon> :p 16:57:33 <anipbu> colindixon: i think oyu have to remove/delete the -1 verify. then do you own +2, then merge. 16:58:07 <colindixon> anipbu: I can't do that 16:58:11 <colindixon> trying recheck to clear it 16:58:43 <colindixon> merged 16:58:54 <lori> colindixon: better avoid recheck 16:59:05 <colindixon> we can kill the jobs 16:59:13 <colindixon> it did remove the -1 verify 16:59:37 <lori> ok 17:00:05 <colindixon> lori: what's next? 17:00:29 <anipbu> I think we can merge all of these: {alto, didm, faas, l2switch, lacp, topoprocessing} 17:00:41 <lori> yes 17:01:02 <anipbu> these should all all be done in parellel since they do not have any immediate dependencies. 17:01:12 <colindixon> working on it 17:01:36 <anipbu> do not merge sdniterfaceapp (this is still pending bgpcep) 17:02:35 <colindixon> doing l2switch first 17:02:56 <colindixon> I have to do recheck because I need jenkins-releng to give up it's -1 17:03:27 <colindixon> unless somebody knows another better way to do it? 17:04:00 <colindixon> l2switch is going 17:04:38 <vrpolak> colindixon: In Reviewers, there is jenkins-releng. Click on x and its votes are gone. 17:04:58 <milfabia> bgpcep done 17:05:00 <colindixon> vrpolak: no x for me b/c i'm not a committer on those projects :-/ 17:05:20 <gvrangan> I am recehcking the vtn patch 17:05:53 <gvrangan> i will merge the vtn patch when verify is done 17:06:26 <colindixon> I'm merging faas now 17:06:28 <lori> vtn needs ovsdb to finish it's merge job 17:06:36 <lori> gvrangan: ^^^ 17:06:50 <colindixon> #info I've merged nemo and l2switch 17:06:53 <lori> gvrangan: so verify will fail before that, I suspect 17:07:02 <vrpolak> colindixon: In that case recheck is probably unavoidable. 17:07:15 <colindixon> #info I've merged faas 17:07:35 <colindixon> alto, didm, lacp, and topoprocessing are up now as they don't have any dependencies 17:07:52 <anipbu> gvrangan: please do not merge vtn until phase 5 17:08:14 <colindixon> anipbu: the 4 above are good though? 17:08:40 <anipbu> colindixon: yes, alto, didm, lacp, and topoprocessing are good to merge 17:09:23 <colindixon> #info lacp merged 17:09:29 <anipbu> bgpcep finished successfully 17:11:02 <anipbu> #info status: merging NEMO, L2SWITCHm FAAS, LACP right now 17:11:29 <colindixon> #Info alto merged 17:12:39 <anipbu> colindixon: sdninterfacapp can be merged as well since bgpcep completed. 17:12:40 <gvrangan> just checked ovsdb was merged... 17:12:43 <gvrangan> okay I will wait 17:13:22 <colindixon> #info didm merged 17:13:37 <colindixon> #info topoprocessing merged 17:14:37 <anipbu> #info Phase Two Complete: merge job complete and artifacts uploaded to nexus for AAA, BGPCEP, NETCONF, NEUTRON, OPENFLOWJAVA, SXP, USC 17:15:33 <colindixon> anipbu: awesome 17:15:47 <lori> gvrangan: yes, but the merge job has to finish, so that artifacts are published to nexus 17:15:54 <colindixon> #info sdninterfaceapp merged 17:16:59 <lori> nemo merge job successful 17:17:10 <lori> ovsdb too 17:17:13 <colindixon> does that unblock anything? 17:18:12 <colindixon> I think sfc is blocking on l2switch, but maybe that's not riht 17:18:15 <anipbu> #info Phase Three Complete: merge job complete and artifacts uploaded to nexus for OFCONFIG, OPENFLOWPLUGIN 17:18:58 <lori> colindixon: no, sfc doesn't block on l2switch 17:19:00 <colindixon> unimgr and vtn should be good to go (waiting for confirmation from lori and anipbu) 17:19:06 <lori> we can now merge {sfc, genius, unimgr, vtn} 17:19:21 <colindixon> OK 17:19:23 <colindixon> I'll start that 17:19:31 <colindixon> gvrangan: you want to do vtn? 17:20:42 <gvrangan> sure Colin, I will merge VTN 17:20:57 <colindixon> #info sfc merged 17:21:23 <anipbu> ovsdb finished successfully: https://jenkins.opendaylight.org/releng/job/ovsdb-merge-boron/230/ 17:22:25 <colindixon> lori, anipbu, zxiiro: we're not waiting for verifies, right? 17:22:33 <lori> no 17:22:36 <anipbu> no 17:22:38 <lori> just merge jobs 17:22:46 <anipbu> we can go ahead and merge 17:22:55 <colindixon> gvrangan: you can verify +1 the vtn job and then click submit then 17:23:09 <colindixon> #info unimgr merged 17:23:21 <colindixon> #info genius merged 17:25:09 <anipbu> still waiting on VTN. it can be merged now. 17:25:43 <colindixon> #info vtn merged 17:25:47 <lori> gvrangan left the channel 17:25:53 <colindixon> we lost gvrangan, so I just did it 17:25:56 <lori> ok 17:26:41 <gvrangan> colindixon: Thanks for merging , My IRC client got disconnected 17:26:56 <colindixon> faas's job passed verify, but not distribution-check, but that's probably fine 17:27:08 <colindixon> faas merged successfully 17:27:40 <colindixon> does faas free up anything? 17:27:49 <colindixon> gbp? 17:27:51 <zxiiro> alright i'm back 17:28:09 * colindixon waves to zxiiro 17:28:18 <lori> no gbp needs netvirt, sfc too 17:28:20 <colindixon> zxiiro: I had to use recheck to clear the verify-1s to merge things 17:28:23 <colindixon> lori: OK 17:28:49 <zxiiro> colindixon: you can also click the x next to the jenkins-releng username 17:29:00 <colindixon> zxiiro: you can, I can't 17:29:17 <zxiiro> huh really? i thought project owners could 17:29:25 <colindixon> groupbasedpolicy:aaa,controller,dlux,faas,honeycomb,mdsal,netconf,neutron,odlparent,openflowplugin,ovsdb,sfc,sxp 17:29:28 <zxiiro> oh wait committers are project owners 17:29:35 <colindixon> lori: doesn't look like gbp depends on netvirt 17:29:37 <colindixon> just sfc 17:29:46 <colindixon> from here: https://jenkins.opendaylight.org/releng/view/autorelease/job/autorelease-project-report-boron/lastSuccessfulBuild/artifact/dependencies.log 17:29:47 <lori> hmm, ok 17:30:10 <anipbu> zxiiro: row 81 will give you an update on the status of each phase: https://docs.google.com/spreadsheets/d/1zImtd764e-hOgJAxoJKl85fxHCPu2agLfqsBtf13zQY/edit#gid=778533050 17:30:20 <zxiiro> anipbu: thanks 17:30:44 <colindixon> lori: that could be wrong, but it's usually not 17:31:03 <colindixon> somebody should check this quickly at some point: https://lists.opendaylight.org/pipermail/release/2016-July/007247.html 17:31:26 <colindixon> atrium having issues with merge that doen't look like random infra issues 17:31:41 <zxiiro> colindixon: looking 17:32:16 <rovarga> zxiiro: probably bad settings.xml or their distributionManagement section is wrong 17:34:52 <colindixon> vtn-merge-boron 17:34:52 <colindixon> genius-merge-boron 17:34:54 <colindixon> unimgr-merge-boron 17:34:55 <colindixon> sfc-merge-boron 17:34:57 <colindixon> sdninterfaceapp-merge-boron 17:34:58 <colindixon> topoprocessing-merge-boron 17:35:00 <colindixon> didm-merge-boron 17:35:01 <colindixon> alto-merge-boron 17:35:02 <colindixon> lacp-merge-boron 17:35:03 <colindixon> those are all queued up 17:35:50 <colindixon> zxiiro: one property of having had to recheck is that there may be verify, validate and distribution check jobs in the pipe that could be killed 17:36:11 <colindixon> l2switch is successfully merged 17:36:19 <zxiiro> yeah let me know if there's jobs you want me to kill 17:36:48 <colindixon> zxiiro: any verify, validate and/or dist-check jobs relatiing to the above 17:36:57 <colindixon> is netvirt ready to go? lori, anipbu 17:37:12 <colindixon> depends on genius 17:37:14 <anipbu> netvirt depends on genius 17:37:29 <lori> yes, we need genius first 17:38:16 <zxiiro> colindixon: next time if you have to recheck use teh trigger "run-sonar" instead 17:38:25 <zxiiro> it will have the same effect but only launch one job 17:38:29 <colindixon> :p 17:38:40 <colindixon> zxiiro: you could just give me moar cow powers :p 17:39:27 <zxiiro> i think we're working on that 17:39:40 <zxiiro> there's been a request to give committers more jenkins power to run manual job, not sure if that includes cancelling 17:39:42 <lori> sfc merge failed 17:39:42 <anipbu> Yay! alto is finally started it's merge build: https://jenkins.opendaylight.org/releng/job/alto-merge-boron/44/ 17:40:09 <colindixon> didm, alto, lacp, topoprocessing, and sdninterfaceapp all just started building 17:40:11 <lori> not a compile issue 17:40:18 <colindixon> lori: looking 17:40:24 <zxiiro> bothers me that atrium only reached out now about their merge issues 17:40:47 <anipbu> sfc failure link to console for easy access: https://jenkins.opendaylight.org/releng/job/sfc-merge-boron/75/console 17:40:52 <colindixon> ssh-agent] Java/JNR ssh-agent 17:40:53 <colindixon> ERROR: [ssh-agent] Could not register BouncyCastle on the remote agent. 17:40:53 <colindixon> hudson.remoting.RequestAbortedException: java.io.IOException: Unexpected termination of the channel 17:40:59 <colindixon> remerging sfc 17:41:02 <colindixon> that looks like random infra 17:41:06 <zxiiro> looks like they haven't had a merge success in a very long time 17:41:15 <lori> triggered another one for sfc 17:41:34 * colindixon doesn't trigger a second 17:41:36 <lori> didm failed too 17:42:02 <lori> and this does seem to be compilation failure 17:42:41 <lori> may need to send another patch 17:42:44 <lori> looking into it 17:42:46 <colindixon> yeah 17:42:48 <colindixon> thanks lori 17:44:49 <colindixon> unimgr and genius are now running merge jobs 17:45:10 <zxiiro> rovarga: colindixon I think it's because their root pom does not have a link back to odlparent so that's why they are missing distribution management section 17:45:29 <colindixon> that makes sense 17:46:21 <lori> it turns out didmm depends on snmp 17:46:38 <lori> so until snmp is merged, didm will be broken 17:47:59 <colindixon> @lori, do we know if the snmp patch works? 17:48:23 <colindixon> if we're 99% sure, I can merge it take the heat if anyone complains 17:48:35 <lori> I can't be 100% sure, since I wasn't able to test locally yet 17:48:44 <lori> let me look at the patch 17:49:11 <lori> it's quite small, so I'm in that 99% 17:49:17 <colindixon> OK 17:49:42 <colindixon> I'm going to merge it unless anyone strenuously objects and I'll take responsibility if anyone complains 17:50:29 <lori> lacp success 17:51:03 <lori> genius compile fails for me locally 17:51:05 <colindixon> going once 17:51:11 <lori> it looks like there was something merged 17:51:17 <colindixon> (for snmp) 17:51:20 <lori> which didn't cause a merge conflict 17:51:26 <lori> I'll be pushing a fix 17:51:32 <colindixon> cool 17:51:33 <lori> I don't object for snmp colindixon 17:51:33 <colindixon> thanks! 17:51:38 <colindixon> how's sfc? 17:52:29 <lori> let me check 17:52:34 <zxiiro> i'm confused why atrium is thirdparty to itself 17:53:10 <lori> sfc: merge running: https://jenkins.opendaylight.org/releng/job/sfc-merge-boron/76/console 17:53:43 <anipbu> We have made a best effort attempt to communicate with SNMP, but due to resource constraint on the SNMP team, we have not heard from them (unresponsive). Unfortunately their patch is blocking multiple projects and I have no objections to merging snmp patch: https://lists.opendaylight.org/pipermail/snmp-dev/2016-July/000297.html 17:53:43 <colindixon> genius merge failure, but we expected hat 17:53:48 <lori> sdninterfaceapp finished successfully 17:55:06 <lori> so indeed the genius merge job failed, here's a fix: https://git.opendaylight.org/gerrit/#/c/41782/ 17:56:34 <colindixon> #link genius fix merged https://git.opendaylight.org/gerrit/#/c/41782/ 17:58:40 <colindixon> zxiiro: only problem with run-sonar is it's going to take for-ev-er for it to run so jenkins-releng will comment and remove the -1 17:59:01 <colindixon> can you remove the -1 so I can merge it? 17:59:02 <colindixon> https://git.opendaylight.org/gerrit/#/c/41065/1 17:59:09 <zxiiro> colindixon: yes 17:59:19 <zxiiro> merged 17:59:33 <colindixon> #info snmp merged 17:59:47 <colindixon> unimgr done 18:00:10 <anipbu> alto completed succesfully: https://jenkins.opendaylight.org/releng/job/alto-merge-boron/44/ 18:02:13 <colindixon> we're down to 5 patches 18:02:24 <colindixon> netvirt, gbp, tsdr, vpnservice, and nic 18:04:54 <anipbu> do we ahve to rerun didm since it's merge job failed? 18:05:51 <lori> unimgr finished successfully 18:06:05 <lori> for didm we need to wait until snmp merge job finishes 18:06:18 <anipbu> lori: okay. 18:06:21 <colindixon> lori: I'd guess so, rihgt? 18:06:22 <lori> also, I discovered an issue similar to genius with didm 18:06:28 <lori> submitting a fix soon 18:07:10 <anipbu> lori: i put didm and tsdr in phase 7 since they both depend on snmp 18:07:52 <zxiiro> lori: can you have a quick look at the atrium compile failure at the bottom and let me know if it's because of the work we're doing right now? 18:07:52 <lori> well, snmp was merged, so it should be ready soon 18:07:54 <zxiiro> https://jenkins.opendaylight.org/releng/job/atrium-verify-boron/3/jdk=openjdk8,nodes=dynamic_verify/console 18:08:01 <zxiiro> I'm assuming it is 18:08:18 <lori> yes, it doesn look like that 18:08:30 <lori> I'll prepare a patch for them too 18:09:04 <zxiiro> lori: just update my existing patch and take over authorship of it and add me as a Co-Authored-By: Thanh Ha <thanh.ha@linuxfoundation.org> at the bottom 18:09:24 <zxiiro> lori: it's patch to fix their merge job anyway 18:10:06 <colindixon> :p apparently atrium needed a fix :p 18:10:16 <zxiiro> lori: or you can do a separate patch and i rebase but they still need both patches to fix their merge job 18:10:31 <lori> here's a fix for didm: https://git.opendaylight.org/gerrit/#/c/41785/ 18:10:39 <lori> should be merged once the snmp merge job passes 18:11:24 <colindixon> snmp and genius merges are sitting in the queue 18:11:30 <colindixon> behind netvirt and didm verifies 18:12:15 <zxiiro> verifies have precedence but if we don't need thedm i can kill them 18:13:28 <lori> netvirt verify can be killed, since it will fail anyway 18:14:02 <lori> didm verify will, too, since snmo didn;t finish 18:15:13 <colindixon> I'm trying :p 18:15:56 <colindixon> do we know why some builders show as (offlline)? 18:19:00 <colindixon> genius and snmp merges are started at least 18:22:22 <zxiiro> colindixon: builders are one-shot now 18:22:35 <zxiiro> colindixon: so they are marked offline for deletion after a job as run on them to guarentee that no other job takes that slave 18:22:53 <zxiiro> sfc finished 18:23:18 <zxiiro> lori: anything we can merge post sfc? 18:24:25 <colindixon> vtn merge failed? 18:25:21 <colindixon> snmp and vtn both timed out 18:25:24 <colindixon> I'm remerging them 18:25:47 <lori> zxiiro: let me check 18:26:23 <zxiiro> grr that connection timout issue is annoying 18:26:32 <colindixon> zxiiro: do we know what it is? 18:27:24 <zxiiro> colindixon: our 2 theories is disk or maybe ldap /cc tykeal 18:27:56 <tykeal> I'm working on the ldap thing... don't have a solution just yet 18:28:14 <colindixon> no worries, OK, those two jobs are waiting on remerge 18:28:17 <tykeal> I'm honestly hoping that's the issue because if it's disk / network related it's going to be harder to troubleshoot, even with Rackspace 18:28:27 <colindixon> should hit soon 18:28:37 <colindixon> startednow 18:28:46 <anipbu> lori: colindixon: Can groupbasedpolicy run now that sfc has finished successfully? https://jenkins.opendaylight.org/releng/view/autorelease/job/autorelease-project-report-beryllium/lastSuccessfulBuild/artifact/dependencies.log 18:29:01 <lori> colindixon: anipbu: not sure about that 18:29:10 <colindixon> I think yes 18:29:21 <colindixon> lori: thinks they rely on netvirt 18:29:54 <colindixon> spectre:groupbasedpolicy ckd$ grep -r netvirt . 18:29:58 <lori> we can try GBP, sure 18:30:05 <colindixon> 80be6ed54c81616bb780330b0afe3280c60bc9eb Merge "Tests for neutron-ovsdb" 18:30:08 <colindixon> so, I think we're good 18:30:30 <colindixon> zxiiro: https://git.opendaylight.org/gerrit/#/c/40829/ 18:30:32 <colindixon> merge it 18:30:47 <zxiiro> done 18:31:11 <colindixon> snmp merged successfully 18:31:18 <colindixon> so didm and tsdr are ready to go 18:31:26 <zxiiro> ok merging... 18:31:54 <anipbu> #info status: merging GROUPBASEDPOLICY right now 18:32:21 <zxiiro> didm and tsdr merged 18:34:08 <colindixon> zxiiro: any idea what's going on that's making this wait on a sonar job that's not coming? 18:34:36 <zxiiro> colindixon: I may have cancelled it which project where you running? 18:34:45 <colindixon> snmp 18:34:50 <zxiiro> colindixon: and did you actually want the result 18:35:16 <colindixon> no, I just wanted it to push the results of the merge 18:35:19 <colindixon> which it seems to not be doing 18:35:23 <colindixon> even though it's done 18:36:01 <zxiiro> like push artifacts? or sonar results? 18:36:16 <anipbu> #info status: merging DIDM right now 18:36:17 <colindixon> no, comment on the gerrit patch that the merge was successful 18:36:21 <anipbu> #info status: merging TSDR right now 18:36:35 <zxiiro> colindixon: oh that's weird because it finished 18:36:39 <zxiiro> https://jenkins.opendaylight.org/releng/view/snmp/job/snmp-merge-boron/lastSuccessfulBuild/console 18:36:55 <colindixon> I know 18:37:10 <colindixon> I think it's waiting for the snmp-sonar job to run, but it never even started because we killed it in the queue 18:37:17 <colindixon> so I think jenkins is just never going to comment 18:37:18 <zxiiro> oh yeah i guess it's stuck then 18:37:27 <zxiiro> that's bizare 18:37:54 <zxiiro> i just issued another run-sonar lets see if it clears it up 18:39:53 <colindixon> is netvirt supposed to be merged: https://git.opendaylight.org/gerrit/#/c/40827/ 18:41:02 <anipbu> netvirt waiting on genius 18:41:24 <colindixon> https://git.opendaylight.org/gerrit/#/c/39880/ 18:41:41 <colindixon> does lispflowmapping need a remerge? 18:41:48 <anipbu> Genius is still building and waiting to merge successully: https://jenkins.opendaylight.org/releng/job/genius-merge-boron/67/ 18:43:15 <lori> colindixon: lispflowmapping is ok 18:43:22 <lori> no need for remerge 18:43:25 <anipbu> concern: topoprocessing typically completes its merge job in 30 minutes or less (#114: 12 minutes; #115: 22 minutes; #116: 23 minutes). However, the latest merge job for topoprocessing is taking over an hour and still has not completed: https://jenkins.opendaylight.org/releng/job/topoprocessing-merge-boron/117/ 18:43:44 <colindixon> didm merge was unstable... 18:43:53 <colindixon> https://jenkins.opendaylight.org/releng/job/didm-merge-boron/40/ 18:44:17 <colindixon> anyone know why? 18:44:39 <lori> looking 18:44:55 <colindixon> anipbu: looks like it's findbugs is going or stalled? 18:45:05 <zxiiro> looks like it's tests causing unstable 18:45:14 <anipbu> it has been sitting on "[FINDBUGS] Computing warning deltas based on reference build #116" for over 30 minutes. 18:45:32 <colindixon> so, I think didm is successful in that it's artifacts are pushed to nexus 18:45:35 <anipbu> proposal: maybe zxiiro can kick off an additional merge job for topoprocessing? 18:45:42 <lori> singlefeaturetest failures 18:46:11 <colindixon> anipbu: I'll remerge topoprocessing 18:46:18 <colindixon> lori: is that for didm? 18:46:25 <lori> colindixon: yes 18:46:54 <colindixon> do we know if it's real? 18:47:18 <colindixon> I'm buidling didm locally 18:48:12 <colindixon> tsdr is successful 18:48:47 <anipbu> #info Phase Three Complete: merge job complete and artifacts uploaded to nexus for ALTO, FAAS, L2SWITCH, LACP, NEMO, OVSDB, SDNINTERFACEAPP 18:49:53 <colindixon> gbp failed 18:50:08 <colindixon> looks compilation related 18:50:26 <colindixon> [ERROR] /w/workspace/groupbasedpolicy-merge-boron/neutron-ovsdb/src/test/java/org/opendaylight/groupbasedpolicy/neutron/ovsdb/OvsdbNodeListenerTest.java:[111,24] incompatible types: org.opendaylight.yang.gen.v1.urn.ietf.params.xml.ns.yang.ietf.inet.types.rev100924.Uri cannot be converted to org.opendaylight.yang.gen.v1.urn.ietf.params.xml.ns.yang.ietf.inet.types.rev130715.Uri 18:50:27 <colindixon> [ERROR] /w/workspace/groupbasedpolicy-merge-boron/neutron-ovsdb/src/test/java/org/opendaylight/groupbasedpolicy/neutron/ovsdb/OvsdbNodeListenerTest.java:[121,31] incompatible types: org.opendaylight.yang.gen.v1.urn.ietf.params.xml.ns.yang.ietf.inet.types.rev100924.IpAddress cannot be converted to org.opendaylight.yang.gen.v1.urn.ietf.params.xml.ns.yang.ietf.inet.types.rev130715.IpAddress 18:50:38 <colindixon> lori ^^^^^^ 18:51:00 <lori> awsome :( 18:51:05 <lori> will take a look 18:51:08 <zxiiro> genius is about to finish 18:51:17 <colindixon> looks like missing depedencies? 18:51:20 <lori> yes 18:51:25 <lori> most likely 18:51:42 <anipbu> zxiiro: it seems topoprocessing build #118 will not start until build #117 ends. And build #117 seems to be hanging. Is there anyway we can kick off #118 while waiting for #117? 18:52:15 <zxiiro> anipbu: it's aborted 18:52:34 <colindixon> I just did that 18:52:45 <lori> colindixon: actually no, there was something else merged since the last rebase :( 18:52:53 <lori> will send a fix for GBP 18:54:37 <zxiiro> genius done, i'll merge netvirt now 18:55:44 <lori> BGP update: https://git.opendaylight.org/gerrit/#/c/41787/ 18:55:50 <lori> I mean GBP 18:56:12 <colindixon> https://git.opendaylight.org/gerrit/41788 18:56:18 <colindixon> I think that will fix didm's instability 18:58:28 <anipbu> #info status: merging NETVIRT right now 19:00:51 <anipbu> vtn finished successfully: https://jenkins.opendaylight.org/releng/job/vtn-merge-boron/119/ 19:02:06 <zxiiro> I think we need to wait for gbp to go any further 19:02:12 <colindixon> can we merge atrium? 19:02:32 <colindixon> I guess we'd need a +2 19:02:37 <colindixon> and it's a leaf so we don't really care 19:02:42 <colindixon> they can fix it when they have time 19:02:49 <anipbu> #info Phase Five Complete: merge job complete and artifacts uploaded to nexus for GENIUS, SFC, UNIMGR, VTN 19:04:26 <lori> colindixon: let's wait for the verify to pass for atrium 19:04:32 <lori> since it has other issues as well 19:04:50 <lori> this is the same patch that switches them over to odlparent 19:04:58 <lori> but it looks like only partially 19:05:07 <lori> since the have a common/ folder 19:05:21 <lori> they are broken anyway 19:05:48 <colindixon> so, we're waiting on netvirt, gbp and topoprocessing now 19:06:29 <anipbu> that is correct. then we can merge nic. then vpnservice. 19:06:58 <anipbu> I recommend not forcing the merge for Atrium, since there projects has other issues beyond the scope of IETF Types Migration Effort. 19:07:51 <colindixon> I think it's really just vpnservice depends on nic depends on gbp as the long pole 19:08:22 <anipbu> *so happy* to see column B slowly turning green over the course of the last six hours 19:09:18 <Prem_> @colindixon VPNService does not have dependency on nic 19:09:44 <colindixon> it does according to this: https://jenkins.opendaylight.org/releng/view/autorelease/job/autorelease-project-report-boron/lastSuccessfulBuild/artifact/dependencies.log 19:10:43 <Prem_> got it, Inocybe team had created Intent over VPNService and since Intent depends on NIC, it is included. 19:10:52 <colindixon> <dependency> 19:10:52 <colindixon> <groupId>org.opendaylight.nic</groupId> 19:10:54 <colindixon> <artifactId>features-nic</artifactId> 19:10:55 <colindixon> <classifier>features</classifier> 19:10:57 <colindixon> <version>${nic.version}</version> 19:10:58 <colindixon> <type>xml</type> 19:10:59 <colindixon> <scope>runtime</scope> 19:11:00 <colindixon> </dependency> 19:11:01 <colindixon> in vpnservice/features/pom.xml 19:11:14 <colindixon> I'm not sure if that's a real dependency 19:11:52 <Prem_> @colindixon VPNIntent was created over VPNService by Inocybe and hence, they have added dependency on NIC 19:12:21 <colindixon> fwiw, I locally verified https://git.opendaylight.org/gerrit/#/c/41788/1 fixes the instability in didm, and it wasn't anything to do with the ieft*types 19:12:29 <Prem_> @colindixon Sorry for the confusion 19:12:37 <colindixon> Prem_: no worries 19:13:31 <anipbu> colindixon: since you're a committer on NIC, can you add a +2. The PTL has already +1 and agreed to the merge (https://lists.opendaylight.org/pipermail/nic-dev/2016-July/001367.html) 19:13:59 <zxiiro> we have a lot of npm projects now... we should probably start looking into proxying the npm registry at some point to speed up builds 19:14:12 <anipbu> colindixon: or if you like, I can ping their PTL again: https://git.opendaylight.org/gerrit/#/c/40825/ 19:14:26 * zxiiro noticed gbp was pulling in npm packages 19:15:36 <anipbu> colindixon: it's just that all other projects have +2 except for NIC, however the TSC has given us permission to merge even with a +1. 19:15:40 <lori> the atrium project passed verify patch https://git.opendaylight.org/gerrit/#/c/41781/ 19:16:47 <lori> but it has no +1 or +2 19:18:01 <colindixon> we'll let them take care of it 19:18:05 <colindixon> it's not blocking others 19:20:57 <lori> sounds good 19:21:49 <anipbu> Atrium has been notified by email and have have updated the spreadsheet accordingly: https://lists.opendaylight.org/pipermail/atrium-dev/2016-July/000012.html 19:22:58 <colindixon> topoprocessing done but unstable 19:23:01 <lori> yeap 19:23:25 <colindixon> building locally to check 19:29:57 <zxiiro> gbp about to finish 19:30:48 <colindixon> yup 19:30:51 <colindixon> looks happy 19:31:24 <colindixon> unstable, but done 19:31:45 <colindixon> checking locally 19:32:58 <colindixon> topoprocessing does appear to have issues 19:33:02 <colindixon> but I don't understand them 19:33:07 <zxiiro> it didn't deploy 19:33:12 <colindixon> PathTranslatorTest.startup:93->createTestContext:87->parseYangSources:326 » SomeModifiersUnresolved 19:33:16 <zxiiro> looks like merge jobs don't deploy unless they are blue 19:33:31 <colindixon> testLegalPathInventoryTp(org.opendaylight.topoprocessing.impl.translator.PathTranslatorTest) Time elapsed: 0.083 sec <<< ERROR! 19:33:32 <colindixon> org.opendaylight.yangtools.yang.parser.spi.meta.SomeModifiersUnresolvedException: Some of SOURCE_LINKAGE modifiers for statements were not resolved. 19:33:33 <colindixon> at org.opendaylight.yangtools.yang.parser.stmt.rfc6020.ImportStatementDefinition$RevisionImport$1.prerequisiteFailed(ImportStatementDefinition.java:171) 19:33:35 <colindixon> at org.opendaylight.yangtools.yang.parser.stmt.reactor.ModifierImpl.failModifier(ModifierImpl.java:91) 19:33:36 <colindixon> at org.opendaylight.yangtools.yang.parser.stmt.reactor.SourceSpecificContext.failModifiers(SourceSpecificContext.java:296) 19:33:38 <colindixon> at org.opendaylight.yangtools.yang.parser.stmt.reactor.BuildGlobalContext.addSourceExceptions(BuildGlobalContext.java:236) 19:33:39 <colindixon> at org.opendaylight.yangtools.yang.parser.stmt.reactor.BuildGlobalContext.completePhaseActions(BuildGlobalContext.java:307) 19:33:41 <colindixon> at org.opendaylight.yangtools.yang.parser.stmt.reactor.BuildGlobalContext.buildEffective(BuildGlobalContext.java:197) 19:33:42 <colindixon> at org.opendaylight.yangtools.yang.parser.stmt.reactor.CrossSourceStatementReactor$BuildAction.buildEffective(CrossSourceStatementReactor.java:135) 19:33:44 <colindixon> at org.opendaylight.topoprocessing.impl.translator.PathTranslatorTest.parseYangSources(PathTranslatorTest.java:326) 19:33:45 <colindixon> at org.opendaylight.topoprocessing.impl.translator.PathTranslatorTest.createTestContext(PathTranslatorTest.java:87) 19:33:46 <colindixon> at org.opendaylight.topoprocessing.impl.translator.PathTranslatorTest.startup(PathTranslatorTest.java:93) 19:36:30 <colindixon> Running org.opendaylight.groupbasedpolicy.resolver.ForwardingResolverTest 19:36:31 <colindixon> Tests run: 4, Failures: 0, Errors: 4, Skipped: 0, Time elapsed: 0.393 sec <<< FAILURE! - in org.opendaylight.groupbasedpolicy.resolver.ForwardingResolverTest 19:36:32 <colindixon> testOnSubtreeModified(org.opendaylight.groupbasedpolicy.resolver.ForwardingResolverTest) Time elapsed: 0.146 sec <<< ERROR! 19:36:33 <colindixon> com.google.common.util.concurrent.UncheckedExecutionException: com.google.common.util.concurrent.UncheckedExecutionException: com.google.common.util.concurrent.UncheckedExecutionException: java.lang.IllegalStateException: Failed to instantiate prototype org.opendaylight.yangtools.binding.data.codec.gen.impl.DataObjectSerializerPrototype as 19:36:34 <colindixon> org.opendaylight.yang.gen.v1.urn.opendaylight.groupbasedpolicy.forwarding.rev160427.forwarding.forwarding.by.tenant.ForwardingContext$StreamWriter 19:36:43 <colindixon> the second burst is GBP 19:38:33 <colindixon> so, do we think we can merge nic? 19:38:48 <lori> colindixon: probably not 19:38:57 <lori> if UNSTABLE merge jobs don't upload 19:39:31 <lori> although nic depends on vtn 19:39:36 <colindixon> zxiiro: do you know? 19:39:37 <lori> and I think not on gbp 19:39:47 <zxiiro> I can modify the job to force UNSTABLE jobs to push artifacts 19:39:51 <zxiiro> if we think we can't fix the tests 19:39:56 <zxiiro> at least it will get us going 19:40:49 <zxiiro> thoughts? 19:40:54 <colindixon> lori: have you looked at the gbp failiure? 19:41:10 <lori> yeah, but it doesn't tell me anything :( 19:41:18 <colindixon> me either 19:41:26 <lori> let me try building locally 19:41:29 <colindixon> it actually looks like a YANG tools bug 19:42:40 <zxiiro> one of didm's failure messages is: No feature named 'odl-openflowplugin-all' with version '0.3.0-SNAPSHOT' available 19:43:06 <zxiiro> i think colindixon already looked at that one 19:43:09 <colindixon> zxiiro: that's fixed 19:43:22 <colindixon> just need to merge this: https://git.opendaylight.org/gerrit/#/c/41788/ 19:43:25 <colindixon> I'll let didm do it 19:43:43 <colindixon> topoprocessing and gbp are the ones that scare me 19:45:25 <lori> I can repro the GBP failure locally 19:46:27 <colindixon> yeah... I just don't know how to even think about starting to debug it 19:47:08 <zxiiro> anyone from gbp here we can pull in? 19:47:13 <colindixon> it does build with -DskipTests 19:48:15 <zxiiro> rerun with skipTests to get us going? and worry about the test failure later? 19:48:43 <lori> zxiiro: yes I think so 19:48:50 <colindixon> zxiiro, lori: I don't know, I think the only people who could debut it are asleep 19:48:57 <colindixon> martin sunal is the obvious person 19:49:00 <lori> I looked at some of the models 19:49:05 <zxiiro> ok let's just skip test it for now 19:49:09 <lori> and I see nothing suspitcious 19:49:17 <colindixon> I think it's a relfection issue 19:49:58 <lori> specifcly forwarding.yang 19:50:00 <anipbu> NIC depends on GBP because nic has GBP Intents 19:51:31 <anipbu> Is it okay for PTL to merge this change? https://git.opendaylight.org/gerrit/#/c/41788/ 19:52:07 <lori> so netvirt merge job failed 19:52:13 <lori> compilation issue this time 19:52:17 <lori> looking into it 19:53:18 <zxiiro> anipbu: I think yes it's up to them 19:55:11 <colindixon> I'm 99% sure that this is part of the reason we're seeing failures in GBP: https://github.com/opendaylight/groupbasedpolicy/blob/master/groupbasedpolicy/src/main/yang/model/policy.yang#L846 19:56:12 <lori> but it has the correct revision 19:56:30 <zxiiro> https://jenkins.opendaylight.org/releng/job/groupbasedpolicy-merge-boron/230/console this is running with skip tests. I'll merge NIC once it passes 19:56:46 <colindixon> lori: my guess is that the tests baked in the old revision 19:56:47 <colindixon> I'm looking 19:57:03 <lori> colindixon: yeah, could be 19:57:51 <lori> this is the netvirt failure: 19:57:53 <lori> [ERROR] Failed to execute goal org.opendaylight.yangtools:yang-maven-plugin:1.0.0-SNAPSHOT:generate-sources (binding) on project natservice-impl: yang-to-sources: Unable to parse yang files from /home/moriarty/src/opendaylight/netvirt/vpnservice/natservice/natservice-impl/src/main/yang: Imported module [ModuleIdentifierImpl{name='odl-interface', namespace=null, revision=2015-03-31, semantic version=0.0.0}] was not found. [at META-INF/yang/ 19:57:53 <lori> natservice-impl.yang:11:4] -> [Help 1] 19:58:16 <lori> the odl-interface.yang that is referenced is a genius artifact 19:58:40 <lori> the artifact is in the dependencies 19:58:50 <lori> so not sure what's up with that 20:00:17 <colindixon> ... 20:03:12 <colindixon> I've got nothing with GBP and I have to leave for a bit 20:03:41 <colindixon> I say that we try to get NIC to build and upload the GBP artifacts to make progress now 20:04:25 <zxiiro> netvirt has been failing since build #142 20:04:26 <zxiiro> https://jenkins.opendaylight.org/releng/view/netvirt/job/netvirt-merge-boron/ 20:05:00 <lori> yes, it looks like genius updated a yang revision 20:05:05 <lori> and netvirt didn't 20:05:11 <lori> i'm testing a fix 20:05:14 <zxiiro> ah ok 20:06:52 <zxiiro> ok gbp artifacts uploading now 20:06:56 <zxiiro> i'm gonna merge nic 20:07:29 <lori> here's the fix for netvirt: https://git.opendaylight.org/gerrit/41789 20:08:15 <lori> zxiiro: did gbp publish artifacts? 20:08:30 <zxiiro> lori: yes it's doing that as we speak 20:08:35 <zxiiro> lori: i ran it with skiptests 20:08:40 <lori> ok 20:14:22 <anipbu> #info status: merging NIC right now 20:16:42 <anipbu> lori: can we say with confidence that GBP unstable build is unrelated to the IETF Type Upgrade? 20:16:57 <lori> anipbu: no, I'm not sure about that 20:17:11 <anipbu> It seems their build is unstable but we went ahead and uploaded their artifacts regardless. 20:17:35 <lori> yes, because they had some downstream dependencies 20:17:43 <lori> and if the code was merged 20:17:52 <lori> it was better to have the artifacts up 20:18:01 <lori> and solve the test failure later 20:18:09 <lori> colin started looking into it 20:18:20 <anipbu> lori: can we say with confidence that GBP unstable build is unrelated to the IETF Type Upgrade or is this still uncertain? 20:18:46 <lori> anipbu: see above 20:19:24 <anipbu> DIDM merged their patch and they are rebuilding here: https://jenkins.opendaylight.org/releng/job/didm-merge-boron/42/ 20:19:59 <lori> great! 20:24:24 <anipbu> lori, am I correct in stating that topoprocessing has unstable build, colin found some topoprocessing error messages related to their tests cases (not compilation error), but we still do not know the root cause. 20:24:43 <lori> anipbu: yes, that's correct 20:28:56 <zxiiro> lori: nic finished, also unstable :( 20:28:57 <zxiiro> https://jenkins.opendaylight.org/releng/job/nic-merge-boron/79/ 20:31:19 <lori> it's singlefeaturetest 20:31:30 <lori> No feature named 'odl-openflowplugin-app-lldp-speaker' with version '0.3.0-SNAPSHOT' available 20:31:48 <zxiiro> yeah looks like it. 20:31:57 <zxiiro> ok i'm gonna skilptest this one too so we can merge the last project 20:32:01 <lori> not sur what it should be replaced with though 20:32:06 <lori> anyone? 20:32:23 <zxiiro> any nic folks here? 20:32:52 <anipbu> lori, am I correct in stating that GROUPBASEDPOLICY has unstable build, colin found some gbp error messages related to their ForwardingResolverTest tests cases (not compilation error), but we still do not know the root cause. We need GBP team to take a look at the error and address the issue. 20:33:32 <lori> anipbu: yes, that's correct 20:33:46 <lori> or any openflowplugin folks? 20:35:58 <anipbu> didm build finished successfully 20:36:06 <anipbu> https://jenkins.opendaylight.org/releng/job/didm-merge-boron/42/ 20:37:07 <lori> excellent 20:40:02 <lori> which project doesn vpnservice depend on? 20:40:55 <zxiiro> nic 20:41:03 <zxiiro> it's our last patch i think 20:41:37 <anipbu> #info Phase Six Complete: merge job complete and artifacts uploaded to nexus for DIDM, SNMP, TSDR 20:41:37 <zxiiro> nic is uploading artifacts now so i'll merge vpnservice 20:43:38 <anipbu> #info status: merging VPNSERVICE right now 20:43:49 <zxiiro> last job here https://jenkins.opendaylight.org/releng/job/vpnservice-merge-boron/118/ 20:49:17 <anipbu> lori, am I correct in stating that NIC has unstable build, you have found some nic error messages related to their singlefeaturetest tests cases (not compilation error). Specifically there seems to be a dependency issue where "no feature named odl-openflowplugin-app-lldp-speaker with version 0.3.0-SNAPSHOT available". Will we expect NIC's team to fix or 20:49:17 <anipbu> are we providing a patch? 20:49:55 <lori> we would need help from openflowplugin to know what feature replaces odl-openflowplugin-app-lldp-speaker, if any 20:50:59 <anipbu> lori: okay. So we expect openflowplugin and nic to work together and provide an appropriate fix for the dependency issue. 20:51:08 <lori> yes 20:53:11 <anipbu> Still waiting on NETVIRT build to complete. Typical wait time is 1 hour 30 minutes: https://jenkins.opendaylight.org/releng/job/netvirt-merge-boron/149/ 20:54:11 <anipbu> Still waiting on VPNSERVICE build to complete. Typical wait time is 50 minutes: https://jenkins.opendaylight.org/releng/job/vpnservice-merge-boron/118/ 20:58:50 <lori> anipbu: I built netvirt locally now, and it passes 20:59:05 <lori> so I expect it to work 20:59:11 <lori> unless we have infra issue 21:04:17 <lori> I'm going mobile, so I might be less responsive 21:04:50 <zxiiro> lori: sounds good, I'm pretty confident we're done for the day anyway 21:06:16 <anipbu> lori: okay. thanks for all the support with regards to the IETF Types Migration Effort 21:19:02 <anipbu> #info Phase Seven Incomplete: pending NETVIRT, GROUPBASEDPOLICY, TOPOPROCESSING. 21:19:09 <anipbu> #info NETVIRT waiting on build to complete in 1 hour 30 minutes: 21:19:14 <anipbu> #link https://jenkins.opendaylight.org/releng/job/netvirt-merge-boron/149/ 21:19:21 <anipbu> #info GROUPBASEDPOLICY build unstable. Error messages related to their ForwardingResolverTest tests cases (not compilation error), but root cause unknown. Kindly ask GROUPBASEDPOLICY team to review issues: 21:19:26 <anipbu> #link https://jenkins.opendaylight.org/releng/job/groupbasedpolicy-merge-boron/228/ 21:19:32 <anipbu> #info TOPOPROCESSING build unstable. Error messages related to their PathTranslatorTest tests cases (not compilation error), but root cause unknown. Kindly ask TOPOPROCESSING team to review issues: 21:19:37 <anipbu> #link https://jenkins.opendaylight.org/releng/job/topoprocessing-merge-boron/118/ 21:19:42 <anipbu> #info Phase Eight Incomplete: pending NIC. 21:19:52 <anipbu> #info NIC build unstable. Error messages related to their singlefeaturetest tests cases (not compilation error). Specifically there seems to be a dependency issue where "no feature named odl-openflowplugin-app-lldp-speaker with version 0.3.0-SNAPSHOT available". Kindly ask OPENFLOWPLUGIN and NIC teams to review issues: 21:19:57 <anipbu> #link https://jenkins.opendaylight.org/releng/job/nic-merge-boron/79/ 21:20:02 <anipbu> #info Phase Nine Incomplete: pending VPNSERVICE. 21:20:07 <anipbu> #info VPNSERVICE waiting on build to complete in 50 minutes: 21:20:12 <anipbu> #link https://jenkins.opendaylight.org/releng/job/vpnservice-merge-boron/118/ 21:22:01 <anipbu> #info For latest status, please review the Google Tracking Spreadsheet: 21:22:06 <anipbu> #link https://docs.google.com/spreadsheets/d/1zImtd764e-hOgJAxoJKl85fxHCPu2agLfqsBtf13zQY/edit#gid=778533050 21:22:15 <anipbu> Thanks to all the folks who joined and helped with the IETF Types Migration Effort IRC Working Session, especially lori, colindixon, zxiiro. 21:22:19 <anipbu> #topic cookies 21:22:23 <anipbu> #endmeeting