16:59:54 #startmeeting tsc 16:59:54 Meeting started Thu May 28 16:59:54 2015 UTC. The chair is colindixon. Information about MeetBot at http://ci.openstack.org/meetbot.html. 16:59:54 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:59:54 The meeting name has been set to 'tsc' 16:59:59 #topic agenda bashing and roll call 17:00:13 #info dlenrow 17:00:32 #info colindixon 17:00:39 #link https://meetings.opendaylight.org/opendaylight-meeting/2015/tsc/opendaylight-meeting-tsc.2015-05-21-16.59.html minutes to previous meeting 17:00:55 #link https://wiki.opendaylight.org/view/TSC:Main#Agenda Agenda for today’s meeting 17:01:10 #info edwarnicke 17:01:30 thanks tbachman 17:01:34 colindixon: np! 17:01:38 * tbachman waits for chair 17:01:51 #action colindixon to try to find somebody to help with documenting the general procedure for the platform upgrade 17:01:54 #chair tbachman 17:01:54 Current chairs: colindixon tbachman 17:01:58 anyone? 17:02:09 bueller? bueler? 17:02:15 edwarnicke: LOL 17:02:35 * tbachman has much catching up to do 17:02:52 #info mohnish anumala 17:03:03 should we save that for infra? 17:03:06 #action LuisGomez and tykeal and jamoluhrsen to work out getting a patch merged in openstack to view all historical performance data 17:03:10 k 17:03:12 4 TSC members sof ar 17:03:28 #info LuisGomez 17:03:33 I’ll get that link 17:03:46 * zxiiro is having issues connecting to webex 17:03:53 #info Chris Price 17:03:57 6 17:04:06 #link https://lists.opendaylight.org/pipermail/release/2015-May/002443.html Email from phrobb on Beryllium release plan proposal 17:04:13 #actoin ChrisPriceAB to work with rovarga, jmedved, et. al. to look into leveraging OPNFV infrastructure for performance measurements and testing 17:04:14 #info Chris Wright 17:04:21 7, gtg 17:04:31 #action ChrisPriceAB to work with rovarga, jmedved, et. al. to look into leveraging OPNFV infrastructure for performance measurements and 17:05:01 Sleep is for the weak 17:05:09 #info jmedved 17:05:11 #topic events 17:05:12 ... Phil is strong :) 17:05:27 #info phrobb says ONS is coming up in a few weeks 17:05:36 #link http://www.opendaylight.org/news/events/ the events page in it’s usually place 17:05:45 * zxiiro is connect with voice only 17:05:56 #info phrobb says they’re looking at shifting some things around for the summit — going to put out an agenda for the developer forum shortly after the Lithium release 17:06:30 #info colindixon asks for additional topics 17:06:48 #info edwarnicke says he’d like to discuss the IPv6 issues with respect to branch-cutting 17:07:43 #topic rovarga says he doesn’t see the restconf/netconf and other creation reviews (breakout projects) 17:07:48 #undo 17:07:48 Removing item from minutes: 17:07:57 #info rovarga says he doesn’t see the restconf/netconf and other creation reviews (breakout projects) 17:08:06 #info colindixon says he’ll schedule this for next week, given the time frame 17:08:19 #topic Lithium and stable/helium updates 17:08:43 #info colindixon asks zxiiro what the status is of autorelease and branch cutting 17:09:00 #info zxiiro says autorelease is failing b/c l2switch is trying to build a distribution that no longer exists 17:09:24 #link https://bugs.opendaylight.org/show_bug.cgi?id=3398 bug against l2switch building old distribution 17:09:42 #info edwarnicke says l2switch is trying to build a hydrogen style distribution, but that distribution no longer exists 17:11:10 #info colindixon asks if downstream projects should wait to bump their versions 17:12:07 #info rovarga says until we perform the bumps, we don’t have visibility into what breakages are present 17:12:49 #info rovarga says we shouldn’t be shipping Lithium until the corresponding patch in yangtools is merged into Lithium 17:13:04 #undo 17:13:05 Removing item from minutes: 17:13:28 #info rovarga says we shouldn’t be shipping Lithium until the corresponding patch in yangtools is merged (cherry-picked) into stable/lithium 17:13:52 #info colindixon asks if downstream projects do bump their versions, will this cause issues using two different versions of yangtools 17:14:01 #info edwarnicke says this will definitely cause issues 17:14:36 #info colindixon says lithium isn’t blocked on this, but the ability to expose bugs in downstream projects is limited 17:15:55 #info edwarnicke says cutting master without version bumps is meaningless 17:16:18 #info edwarnicke also highlights that a lot of downstream projects feel pressure to "cut" 17:16:34 alagalah: thx! 17:16:47 lots of noise 17:16:55 better, thank you 17:16:56 cdub: Oh fine, I make one comment. 17:17:01 sheesh 17:17:06 lol 17:17:09 heh 17:17:51 #link https://git.opendaylight.org/gerrit/#/c/21268/ <- l2switch stable/lithium should not be blocking autorelease anymore 17:17:53 ok, as a mere human, i am hearing there is (a) no requirement to create stable/lithium and (b) we should not version bump 17:17:57 #info The view of the TSC at this point is that the stable/lithium branch cutting does not block the ability to cut RC releases 17:18:01 colindixon: does that look good? 17:18:12 dbainbri: That does not match my understanding of what was said 17:18:17 dbainbri: no i don't hthink so 17:18:36 edwarnicke: that's great thanks! 17:19:03 thanks to evanz rather 17:19:08 #info dbainbri asks if there is (a) no requirement to create stable/lithium and (b) we should not version bump 17:19:16 zxiiro: Let the record reflect that it merged 8 minutes ago, and so you were 100% factually correct in what you said when you said it :) 17:19:54 #info colindixon says there is currently no value in cutting stable/lithium unless you bump on master; if you cannot bump all your versions on master, you should not cut stable/litihium b/c you get no value, but does increase complexity 17:20:03 colindixon: You did not repeat what I said, you said a version of it that is clearer ;) 17:20:33 * tbachman didn’t quite get that last part 17:20:58 #info colindixon says that not bumping master, you have two things publishing artifacts to nexus, with no way of saying which one ‘wins’ 17:21:16 #info looking at OFP master, I see ./parent/pom.xml: 0.8.0-SNAPSHOT 17:22:12 #action zxiiro to send email to release to asking projects to not cut stable/lithium without version bumping 17:22:20 oops 17:22:21 rovarga: So it does appear that OFpluing master depends on the correct yangtools version then? 17:22:23 poor grammar there 17:22:29 #undo 17:22:29 Removing item from minutes: 17:22:47 #action zxiiro and gzhao to send email to release asking projects to not cut stable/lithium without version bumping 17:23:04 #info ebrjohn says that SFC has already cut stable/lithium and not bumped versions 17:23:18 #info colindixon says you technically shouldn’t merge to master w/o the version bump 17:23:34 edwarnicke: yup 17:23:52 #info edwarnicke says it may be easier for SFC to delete the current stable/lithium branch, and then redo it when you’re ready to version bump 17:24:17 tbachman: thanks, I have IRC only right now 17:24:24 gzhao: np! 17:24:24 edwarnicke rovarga - I believe we are pointing to correct Yangtools version with your patch https://git.opendaylight.org/gerrit/#/c/20316/ 17:25:09 #link https://git.opendaylight.org/gerrit/#/c/20316/ patch from edwarnicke that bumped yangtools version in openflowplugin 17:25:20 * edwarnicke apparently can't even keep track of what he himself has done 17:25:34 #info LuisGomez says that integration already did their branch — without the branch, there’s no way they can do their testing 17:26:00 * tbachman loves the smell of integration in the morning 17:26:12 s/integration/branch-cutting/ 17:26:33 where’s regXboi when you need him to get a movie reference? 17:26:42 smells like victory 17:26:46 lol 17:26:50 * cdub helps tbachman out 17:26:51 teen spirit even? 17:26:53 lol 17:26:56 mvn dependency:list ;) 17:26:59 17:27:27 * tbachman is glad that cdub got the ref 17:27:46 https://wiki.opendaylight.org/view/Simultaneous_Release:Lithium_Release_Plan#Project_Dependency_Diagram 17:27:53 #info edwarnicke says there are a lot of projects that depend on OVSDB, so it’s probably the next project that needs to cut and bump 17:27:54 tbachman: but what about the inference ? 17:28:01 movie 17:28:05 think war movie 17:28:07 tbachman: yes, and ... 17:28:09 #link https://jenkins.opendaylight.org/releng/view/autorelease/job/autorelease-daily-master/lastFailedBuild/artifact/dependencies.log 17:28:10 tbachman: inference... 17:28:16 reference 17:28:18 "napalm == integration" ?? 17:28:22 tbachman: sigh 17:28:27 branch-cutting 17:28:28 had to fix it 17:28:34 zxiiro: thx 17:28:58 rovarga: is that you? 17:29:09 lots of noise again 17:29:09 I am trying to get to talk :) 17:29:48 ebrjohn: so are you suggesting that rovarga talking is "lots of noise" … :) 17:29:55 rovarga: Going off mute seems to be a successful strategy for you, as the squelch is just amazing... 17:30:33 #action zxiiro and gzhao to work to drive the version bump through projects using the dependency chart as a guide: it seems, like the next targets would be OVSDB and L2 Switch, followed by dlux 17:31:52 isn’t the big concern here integration? 17:32:08 #info rovarga suggests that we force merge the version bump on master and then deal with the issue later to return the build to normal, edwarnicke sees dragons there 17:32:09 the offset 0 and 1 projects are all cut and bumped 17:32:24 tbachman: but there are dependendencies within offset 2 17:32:51 colindixon: yeah, but my point being that the main thing depending on offset 2 projects is integration 17:32:57 colindixon: did you see that email I sent today about problems with the Vesion-bump.sh script? 17:32:58 tbachman: yes 17:32:59 and that’s our “health meter” 17:33:15 * ChrisPriceAB appreciates Ed's social consciousness. :O 17:33:16 ebrjohn: I didn’t, but I’ll look later 17:33:23 edwarnicke: fair enough 17:33:49 ChrisPriceAB: I have scars, so many scars ;) 17:33:56 ;-) 17:34:11 crazy talk 17:34:21 #link https://docs.google.com/spreadsheets/d/1KPpO9LH539Vlcoa4RvLa6PPCdLifi5JD-ihRhlybqeo/edit#gid=676729675 blocking bugs 17:34:29 colindixon: edwarnicke: it is an option, I think we should try another week and see whether we actually in sucha bind where it is useful :) 17:34:39 #link https://wiki.opendaylight.org/view/Weather link to weather map page 17:35:01 This branch cutting / version bump has been very painful this time around :) 17:35:59 #topic System Integration and Test 17:36:05 edwarnicke: and the titanic cross went smoothly, except for that iceberg thing 17:36:15 #info OVS 2.3 is working in CI, yay! 17:36:17 hehe 17:36:36 LuisGomez: What are we doing around OVS2.4 ? 17:36:43 #info LuisGomez says they cut stable/lithium yesterday 17:37:12 LuisGomez: Its not released but will be soon... just like Lithium :) 17:37:22 #info LuisGomez says their main issue is that all their triggers are “weird” now, so when something fails, it’s hard to know whether it was from stable/lithium or from master 17:37:44 #info the key issue is that cross-project triggers are still intra-branch, e.g., master triggerst master, stable/lithium triggers stable/lithium, but that’s not the reality on the ground at the moment 17:39:38 btw. who do I talk to to get BUG-3051 on the blocker list? :) 17:40:22 * rovarga will follow-up :) 17:40:27 #action colindixon to follow up with zxiiro and gzhao about version bumps on friday and monday to see if we need interim solutions 17:40:43 #action colindixon to add bug 3051 the the blocker list and get robert edit access 17:40:46 colindixon: thx! 17:41:00 #topic Infrastructure 17:41:29 #info tykeal says last night there were issues with slaves not working right — resulting in a huge queue today 17:41:43 #info colindixon asks how long until this this resolves itself 17:41:51 jenkins is building as fast as it can ;) 17:41:54 #info tykeal says maybe by midnight — but hard to know for sure 17:42:00 run jenkins, run! 17:42:56 #info rovarga asks if we could expand the support staff so that we could reach peopel for issues 24x7 17:43:24 #action colindixon to take the issue of more staff for added hours to board 17:44:09 #link https://lists.opendaylight.org/pipermail/tsc/2015-May/003109.html email from ttkacik about LF infra 17:44:34 * ChrisPriceAB thinks no news is good news... :s 17:44:37 #topic berullium release plann 17:44:43 #undo 17:44:43 Removing item from minutes: 17:44:44 #topic Beryllium Release Planning and Board Requests 17:44:52 #link https://wiki.opendaylight.org/view/Simultaneous_Release:Beryllium_Release_Plan the draft release plan 17:45:13 #info colindixon asks people to look over it and comment on mailing lists 17:45:36 eek, mid January... 17:45:46 #info colindixon says there’s also an issue that some of the document may not reflect the project lifecycle changes 17:45:56 #action colindixon to draft changes to language around project lifecycles 17:46:29 #info edwarnicke says he feels very bad about voting on that by next Wednesday, as he’s been heads down in development mode 17:46:39 :) 17:47:29 Did someone say JIRA 17:47:31 ??? 17:47:32 :) 17:47:34 #topic Moving to Jira 17:48:07 #link https://lists.opendaylight.org/pipermail/tsc/2015-February/002554.html mailing list thread started here 17:48:07 #info rovarga says he’d like to get feedback from the community about using bugzilla vs. Trello vs. other tools 17:48:13 #link https://lists.opendaylight.org/pipermail/tsc/2015-March/002730.html and moved here 17:48:18 colindixon: thx! 17:48:21 #link https://lists.opendaylight.org/pipermail/tsc/2015-May/003106.html before ending here 17:48:25 :) 17:49:26 #info alagalah would use JIRA for GBP if ODL got a FOSS license from Atlassian 17:49:37 #info rovarga says this discussion has been brought up on the mailing lists 2 or 3 times, and has stalled each time; would like to get closure on this — worth moving to Jira? Too much of a disturbance, etc.? 17:49:43 * ChrisPriceAB has been working in Jira recently. Finds it quote useful and can help with the project activity planning and tracking. 17:49:57 * colindixon notes that his OF plugin build (kicked off when I found out that yangtools version was also bumped) just succeeded, so yay 17:49:57 #info rovarga also would like to know how the TSC and PTLs feel about using Jira 17:50:29 #info edwarnicke says we can split this into two things: can we switch from bugzilla to jira; the other is: how do we manage migrating from bugzilla to jira 17:50:47 #info colindixon says there’s another piece — running and operating jira 17:51:05 #info colindixon says there’s a cost to migrate from bugzilla to jira, and an operational cost for running jira 17:51:22 Do we have 4 Questions? going once... going twice... Thanks Ed! 17:51:33 #info edwarnicke says that in the absence of the cost, there’s the question of using jira vs. bugzilla 17:51:37 ChrisPriceAB: lol! 17:51:43 * edwarnicke likes Jira :) 17:51:44 ChrisPriceAB: Is that a 5th question ? <<<---- 6 17:51:48 why are you surprised colindixon that the OF plugin build succeeded? :) 17:51:49 lol 17:52:41 #info colindixon says that his perception from the mailing list was that people preferrred Jira, but there’s a big cost 17:52:58 * ChrisPriceAB likes the cross project tracking in Jira. 17:53:09 #info colindixon asks that cost aside, are folks opposed to moving to Jira 17:53:20 #info alagalah asks if we’d adopt it to everything, or just use it for bug-tracking 17:53:44 #info rovarga says we could start with bug-tracking, and add features as the community becomes more comfortable with it 17:54:18 #info ChrisPriceAB recommends using Jira for planning as well 17:54:34 #info colindixon says that his understanding is that not all plugins are available for opensource 17:54:51 lol edwarnicke 17:55:07 there should be no cost 17:55:20 atlassian is free for opensource/non-profits 17:55:26 snoble: all of it? 17:55:32 no not all of it 17:55:32 yes 17:55:41 #info abhijitkumbhare says that most people have used bugzilla but not used Jira 17:55:43 most of it 17:55:48 ok, maybe tykeal knows more about it 17:55:52 The difficulties of Bugzilla are related to scalability as you get more projects/projects, versions, workflows 17:56:05 we use plugins from other companies that have cost but we get them free 17:56:09 * tykeal notes we run several jira instances for various projects 17:56:29 #info dneary says that bugzilla has difficulties related to scalability as you get more projects, versions, workflows 17:56:37 #info tykeal notes that they run several jira instances for various projects 17:56:53 ChrisPriceAB, I'm not a fan of Jira in general (personal opinion, not based on a huge amount of recent experience) 17:56:54 oh gosh no, I didn't mean for that to become an info! 17:56:55 :P 17:57:03 #info rovarga recommends setting up some sort of sandbox where people can try out how their workflows would be 17:57:07 tykeal: lol 17:57:08 #action phrobb to investigate what it takes to get to the license, and what the licence would be for 17:57:19 tykeal: you are an authority ;) 17:57:26 * tykeal cries about that 17:57:28 lol 17:57:30 hehe dneary, I'm not a fan of any tool I have tried in this space tbh. 17:57:46 #info What tykeal meant is that dneary said that most of Bugzilla complaints are related to the UI failing to scale well to many projects, versions, complex workflows, etc 17:57:56 hehe 17:58:04 rovarga: find a small working group to focus, pull project leaders to review, etc... 17:58:07 So my question was: how many projects, issues, versions are we talking about? 17:58:10 #info we’d like to set up a sandbox and maybe do a TWS call where we show it off 17:58:12 something like that 17:58:28 * tbachman can’t keep up 17:58:55 * ChrisPriceAB thinks we should involve aricg from the OPNFV team who is driving an instance today 17:59:14 #endmeeting