17:04:16 #startmeeting 17:04:16 Meeting started Tue Jan 28 17:04:16 2014 UTC. The chair is cdub. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:04:16 Useful Commands: #action #agreed #help #info #idea #link #topic. 17:04:23 #topic roll call 17:04:26 #info Andrew Grimberg for infrastructure support 17:05:18 #info regXboi for openDove (still) 17:06:07 #info Phil Robb here, download page and other 17:06:08 what, you didn't change projects after yesterday regXboi? 17:06:42 #include Chris Wright for ... 17:06:45 * regXboi has thought about it, but who'd have me? 17:06:48 heh, *sigh* 17:06:49 #info Abhijit Kumbhare OpenFlow plugin 17:06:53 #info Chris Wright for ... 17:07:17 cdub: is having a problem figuring out what kind of silliness he's here representing today 17:07:52 #info Madhu here too 17:07:55 cdub: I think you're here for all the sleepless ones 17:07:55 tykeal: and having language trouble too 17:08:07 tykeal: heh, that's about right 17:08:12 so... I have a quesiton about the non-osgi things that have to go into the support VMs 17:08:25 #topic artifact release status 17:08:40 so the only one i'm not up to speed on is integration 17:09:02 we had all others done AFAIK when I left last night 17:09:05 any updates? 17:09:23 from the email I saw to discuss we've got artifacts out for integration and there was an ask for folks to do some testing against it 17:09:34 and, any issues from testing? 17:09:49 so... I'll ask again 17:09:58 what about non OSGI artifacts? 17:10:19 shague: are you aware of that? 17:10:28 the only one that I've been seeing is here: https://lists.opendaylight.org/pipermail/discuss/2014-January/001192.html 17:10:51 #info networkstatic static with ovsdb 17:10:53 regXboi: i know your specific interest is in opendove agent, etc 17:10:57 tykeal: yes - I plan on pulling the virt edition and test driving the opendove portion later this morning 17:11:02 I don't know if a bug has been raised about that issue 17:11:10 cdub: integration buids, yes, 17:12:02 shague: meaning there are binary artifacts for opendove? 17:12:43 yeah, anees has been working on the rpm's for it. 17:12:47 cdub: kyle tested the virt edition for ovsdb + devstack integration for OF13 and it is a go. 17:13:04 my understanding from ashaikh was that he had submitted specs for rpms to go into the support VM 17:13:19 and I'm trying to push that across the finish line as well 17:13:58 regXboi: and you work with shague on finishing that? 17:14:02 s/and/can/ 17:14:16 Madhu: cool, #info? 17:14:20 that's why I'm here 17:14:20 * tykeal forgives cdub's lacking of typing skills this morning ;) 17:14:33 tykeal: thanks ;) 17:14:53 regXboi: ok, let's consider you and sam connected and then we can move on 17:14:56 cdub: yes, we have been reviewing the specs. Annes pushed a new one that I am reviewing today 17:15:11 #action regXboi and shague will work to get opendove rpms into VM 17:15:16 shague: please include me in the reply 17:15:22 shague: ok, thanks 17:15:34 #info Kyle confirmed that the released Virt edition (with ovsdb & OF13) integrates fine with devstack 17:15:59 #info initial tests of opendove from virt edition looks good will have more details later 17:16:32 i have an open question which is about 3rd party code in controller project 17:16:49 i'll send to list, but perhaps folks here have a thought? 17:17:02 do we have a list of 3rd party code? 17:17:07 the 3rd party code was runnin as a snapshot for quite some time 17:17:24 and, in particular, openflowj, had to be down-rev'd in snmp4sdn 17:17:54 makes me wonder if we should've released those projects too 17:18:17 regXboi: jersey-servlet org.apache.catalina.filters.CorsFilter com.siemens.ct.exi net.sf.jung2 ganymed openflowj 17:18:50 #info who owns controller/third-party and when do we release those artifacts? 17:19:02 ah yes, thanks 17:19:08 cdub: shouldn't that be more of a help? 17:19:22 fair 17:19:33 #help who owns controller/third-party and when do we release those artifacts? 17:20:59 i'll take silence as no great insights...fair enough ;) 17:21:10 I think EdWarnicke has been tracking those for controller. I've been working with him on making sure source is available for the LGPL component dependencies 17:21:11 heh 17:22:09 phrobb: ok, mainly thiking we have some changes that we might not have picked up by reverting to the last known released version 17:22:11 cdub: 17:22:22 #topic documentation 17:22:43 edwarnicke! you have some thoughts on that? 17:22:52 cdub: Back in September GiovanniMeo_Away cut a trial run release of controller, and cut all the third party except for ganymed 17:23:07 He then commented them out, so they haven't been built or changed since then 17:23:13 So we did not need to respin them 17:23:22 We did respin ganymed 17:23:35 edwarnicke: ok, openflowj had some minor changes which we simply dropped 17:23:36 cdub: Thank you for your attention to detail though :) Please don't ever stop :) 17:23:46 cdub: *sigh* 17:24:00 s/1.6/1.7 java src/target 17:24:04 (additional note: nerves fried today, so if I don't make sense, please call me on it) 17:24:15 and some other minor pom thing (maven plugin maybe) 17:24:26 What were the changes/ 17:24:28 ? 17:25:07 edwarnicke: git show 80fa93b16fe26d6f0d5ea71ff4f387a48b09b5fa (if you're in controller repo) 17:25:50 #link http://fpaste.org/72446/92994413/ 17:26:36 again, not critical so much as something i think we simply "left behind" and should make sure we keep on those things (and I only looked at one or two other thrid-party bits last night before falling over) 17:27:06 so...documentation 17:27:14 anyone here from the Docs team? 17:27:42 we're in the final mile here folks..working code is only a part of the picture 17:28:37 can anyone #info a docs status (realizing it's not likely any different from last friday) 17:29:12 #info For the openflow plugin project - we think the documentation will need end of the week. 17:29:34 abhijitkumbhare: thank you 17:29:41 #info same for opendove 17:29:45 heh 17:29:55 working on it as we sit here 17:30:05 but review time needs to be factored in 17:30:09 * edwarnicke expects to be documenting furiously until the end of the week 17:30:14 regXboi: good point 17:30:17 #info for lisp, same here, we'll have documentation by end of this week 17:30:34 #info documenation is not just writing, but also reviewing and trying the docs 17:30:49 cdub: I think if it as testing the docs... but concur :) 17:31:07 cdub: that's how I'm writing the docs - on the fly as I verify that stuff works :) 17:31:08 #info trying aka testing the docs 17:31:13 regXboi: ok 17:31:36 regXboi: Its not just making sure the thing they describe works, but also making sure that they make sense to another pair of eyes 17:31:53 * edwarnicke often epicly fails at the second test 17:32:03 i'm going to go out on a limb here and say docs are still in the at-risk category... 17:32:10 yes 17:32:13 yes 17:32:25 edwarnicke: hence my comment about review 17:32:29 what is drop dead for doc btw? 17:32:43 #help Documentation needs help. Code isn't usable if it's either undocumented or the docs are unintelligible...Please Help! 17:33:16 we need a target :) 17:33:17 networkstatic: i think tomorrow is current plan 17:33:31 roger sir 17:33:33 yikes 17:33:43 cdub: I would actually suggest that may be to soon for most folks. I know tomorrow was the goal 17:33:47 networkstatic: yeah, yikes 17:34:10 maybe a status review 17:34:15 keep pressure up :) 17:34:20 networkstatic: definitely a status review 17:34:40 Have folks started filling in their Release Review templates so that we can hold release reviews/ 17:34:43 ? 17:34:47 cdub: do you mean that tomorrow is too soon for release reviews? 17:35:18 edwarnicke: I sent an email to discuss with a strawman proposal for release reviews 17:36:10 dmm: Thank you, apoligies for being a bit behind this morning 17:36:17 no worries 17:36:45 #info pinged Rob, asked for status and suggeste we consider a realistic deadline if we are off schedule 17:37:08 dmm: well, good question, is docs part of that (sorry, no longer recall what's in there)? 17:37:28 dmm: If memory serves, it is 17:37:42 ok then, i'd say...yup gonna slip 17:38:06 dmm: what pieces to we need settled to do release reviews? 17:38:35 dmm: we've got mechanics of artifact release behind us...testing pending, docs incomplete 17:38:49 https://wiki.opendaylight.org/view/Sample_User_Guide 17:39:25 networkstatic: thnx 17:39:30 sorry that was just an informational link for meeting, wasn't sure if i should summit to notes or note 17:39:33 er not 17:40:21 cdub: right, but the calendar to get this (release reviews) looks challenging 17:40:35 networkstatic: if you #link it gets into the meeting summary nicely (just need some context for minutes readers) 17:40:41 #link https://nexus.opendaylight.org/content/repositories/opendaylight.snapshot/org/opendaylight/integration/ release editions are here. I suspect these are what folks should use for any place in their docs they are referring to the release editions 17:40:53 (which is why I wanted to use the TSC call for this in addition to some other time, TBD) 17:42:04 edwarnicke: reasonble point...part of being cohesive/coherent... 17:42:35 dmm: ok, i'm at a loss for how we'll complete this 17:43:05 dmm: i'd be happy w/ using these IRC meetings for review...which should mainly end w/ a bunch of +1s 17:43:13 #link Sample user guide that each project has to have for their project https://wiki.opendaylight.org/view/Sample_User_Guide 17:43:13 On release reviews, lets start simple 17:43:27 Could folks just get current versions of where they are into their release review docs 17:43:33 here, let's change topic 17:43:38 #topic release reviews 17:44:06 cdub: also would work, as long as we can coordinate; we could do Thursday/Friday 17:44:21 I estimated 15 min/per; realistic? 17:44:45 depends, we've been konwn to digress ;) 17:44:56 cdub: :-) 17:45:23 edwarnicke: can you #info that, and then double down and repeat on the list? 17:46:06 edwarnicke: definitely 17:46:32 dmm: but yeah, something like that...ideally it'd be only 5min, but 15min is reasonable for first try at this 17:46:33 I guess I need to modify my strawman 17:46:51 #info Let's start simple, could folks just get current versions of where they are into their release review docs 17:46:57 #action dmm will update strawman release reviews 17:47:17 #link https://wiki.opendaylight.org/view/Sample_Release_Review - Release Review template 17:47:35 ok anything else we can accomplish on this topic? 17:47:45 Everybody good with asking folks to hang their release review off their project pages? 17:47:59 cdub: probably not other than folks making sure that folks fill out the template 17:48:06 move along... 17:48:30 edwarnicke: wfm 17:49:09 I suggest changing the "Architectural Issues" and "Security Issues" to just "Known Limitations and Bugs" in release review template? 17:49:11 (i do like a central point of indirection so we don't have to chase down 15links, but that's pretty minor) 17:49:12 #info please put release review documents under your project page as 'Hydrogen Release Review for $PROJECT' 17:49:25 all: need to run to another call in 10 minutes 17:49:30 abhijitkumbhare: Could you explain the thinking behind that? 17:49:46 Actually limitations only - bugs is already in bugzilla 17:50:14 regXboi: ok, we're nearing the end here. next up is downloads page and double back on testing now that LuisGomez is here 17:50:23 First of all the limitations are not captured under arch & security issues 17:50:38 alrighty, I'll try and catch up on the minutes later 17:50:53 yes, i am here 17:50:55 regXboi: cool,thanks 17:51:31 dmm: you see abhijitkumbhare toughts on template? 17:51:42 cdub: yeah 17:52:23 abhijitkumbhare: regardless of the title, it's good to know what is known not to work, and it's good to know "don't face this towards the internet" 17:52:40 Yes 17:53:13 Hmm... wouldn't they be captured under the Bug section 17:53:23 abhijitkumbhare: perhaps your suggestion is to just collapse to one, and that the limiation or known issues are self-explanatory? 17:53:30 Yes 17:53:46 I do think we should have the current sections, but an not averse to a 'Known Limitations' section 17:53:46 I mean yes to cdub 17:54:07 (Security issues are often called out as seperate thing) 17:54:27 i think we want to make sure ... ^^ what edwarnicke said 17:54:46 but aside of that, i'm not picky about wording, location, etc... 17:55:56 ok, let's take this to the list as part of the strawman thread to work out specific details 17:56:00 edwarnicke (but security issues may be hard to identify at the moment - so it may be hard to give us a cerificate that there are no known security issues :-) ) 17:56:18 I meant give ourselves a cert 17:56:47 maybe wording can be useful...'security considerations' instead of 'issues' 17:56:58 OK 17:57:11 cdub: yes (security considerations) 17:57:15 alright, let's move on, and finish above on the list 17:57:28 #topic downloads 17:57:49 phrobb: any input here? 17:58:13 shague, tykeal you too? 17:58:55 cdub: there's been some mail flying around with pics of mock-ups for the page ;) 17:59:35 #link https://lists.opendaylight.org/pipermail/tsc/2014-January/000805.html downloads mockups 17:59:56 tykeal: there's one... 18:00:07 cdub: sorry, multi-tasking. Mock up pages sent to tsc Plrease respond to that with your preference. Shague and LF folks working together on what will go where 18:00:10 #info input welcome 18:00:20 some dicsussion around how to add the additional downlaod adrtifacts like D4A and oepndove. 18:00:52 phrobb: thanks 18:01:23 shague: ok, and that discussion is reasonably well in-hand...nothing major that needs group input, right? 18:02:00 yeah, no extra input is needed. 18:02:03 cool, thanks 18:02:07 #topics testing 18:02:13 LuisGomez: you're up 18:02:21 my turn then 18:02:29 we got some initial feedback earlier in meeting...waht's the latest? 18:02:46 (ovsdb and opendove reported that initial testing was positive) 18:03:06 system test passed for the editions built from project releases just before integration release cut 18:03:28 today we will pass the test to the releases 18:03:55 these are in nexus and current automation only fetch from jenkins 18:04:13 thats why we could not test the release artifacts yesterday 18:04:14 ah, how much work to update scripts? 18:04:53 i will just need to create around 8 jobs in jenkins to fetch the releases and test them 18:05:07 like couple of hours maximum 18:05:25 so doable today and results tomorrow? 18:05:39 or i can download the editions and test in ,y local env, that will take less than 30 min 18:05:53 * cdub mumbles automate... ;) 18:06:07 but it is good to automate 18:06:47 we have already a suite for of13 in base edition 18:06:49 ok, so sounds like tomorrow we should have test results either way? 18:06:54 ccccccbugiknnkkeefrkfebhrtrhdhcldrdjelcifbtb 18:06:54 yes 18:07:31 cool (sorry about the garbage above) 18:07:57 #info automating tests to pull release artifacts, expecting results tomorrow 18:08:10 LuisGomez: anything else? 18:08:21 i can also pass vtn mgr system test with of13 option 18:08:37 oh, that's new 18:08:45 affinity system test is still ongoing work 18:09:11 LuisGomez: we had 2 issues w/ of13...one was it didn't work w/ mininet1.0, other was it iddn't work w/ vtn 18:09:14 yes vtn mgr system test is ready since last week 18:09:35 do you mean pass the test, or pass of13 option to the test? 18:09:41 mininet 1.0 got much better 18:10:00 LuisGomez: great that VTN passes with the of13 option 18:10:01 there was a huge issue last week and this week seems fixed 18:10:15 for VTN i do not know, i need to test 18:10:23 ok, that's great, can you #info the update? 18:10:49 the big issue was observed running NSF test with mininrt 1.0 18:10:54 but that one is good now 18:11:01 ok 18:11:27 #info today we will work on automate release test 18:11:44 #info release system test result available today 18:12:01 alright, thanks 18:12:19 let's move on to the final topic 18:12:31 #topic any other topic? 18:13:03 i have nothing...will leave it open for a minute 18:13:12 * cdub sleep 60 & 18:14:22 ok, i think we're done 18:14:26 thanks all 18:14:29 #endmeeting