17:01:19 #startmeeting 17:01:19 Meeting started Fri Jan 24 17:01:19 2014 UTC. The chair is colindixon. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:01:19 Useful Commands: #action #agreed #help #info #idea #link #topic. 17:01:39 #info Prasanna for openflowplugin 17:01:45 #topic roll call and agenda 17:02:09 #info Prasanna for openflowplugin 17:02:11 hey prasanna, Madhu is looking for help in the #opendaylight-openflowplugin channel, just fyi 17:02:12 #info oflibMichal for the openflowjava 17:02:31 #info Phil Robb 17:02:33 #link https://docs.google.com/spreadsheet/ccc?key=0AoSzir1BfjyWdDQyVElWNG9mcWxhblREckZjbjFxUVE#gid=1 the big spreadsheet 17:02:50 ok 17:03:00 #info Anees / opendove 17:03:01 #link https://wiki.opendaylight.org/view/CrossProject:Hydrogen_Release_Work the wiki page for things 17:03:02 #info rovarga for yangtools/bgpcep 17:03:09 #info Chris Wright for making of trouble 17:03:27 edwarnicke, you around? 17:03:42 #info shague download artifacts 17:03:50 #info Madhu ovsdb 17:04:18 #info regXboi from OpenDove is 50% here 17:04:50 so, basically, at this point, we're so close to the final cutting of artifacts that I'm going to have an agenda that's mostly around telling people what per-project code cleanup is done/needs work, a bit about documentation, a bit about OF1.3 and other edition testing, and then open it up 17:04:53 #info Ed Warnicke for controller 17:05:17 #topic per-project code cleanup activities 17:06:05 #info I created a "Per-proj Code Cleanup Summary" tab in the big spreadsheet (see link above) which is filled out based purely on info that was in the project's tab 17:06:19 please look at it 17:06:37 update it if it's wrong 17:06:40 add other tasks 17:06:48 if we need them 17:06:50 #info Abhijit Kumbhare OF-plugin 17:07:19 #info michal_rehak / openflowplugin 17:07:54 I have to check, but I'm pretty sure opendove is finished with licensing... 17:08:02 will verify later today 17:08:16 #info most projects look pretty good, affinity is an obvious exception (not trying to blame, but just assessing) 17:08:41 #action everyone go look at this spreadsheet, update it if it's wrong, and use it as a todo list 17:08:42 #info Andrew Grimberg for infrastructure support 17:09:01 cdub, edwarnicke, Madhu, phrobb: as people who have been wrangling some of these tasks 17:09:06 do you have anything you want to add? 17:09:26 no colindixon the spreadsheet looks good 17:09:44 Better than good (in terms of format) :) 17:09:56 #info also, please feel free to extend it as a kind of dashboard for the overall release status 17:10:16 do we have anyone from affinity on right now? 17:11:05 I'll take that as a no 17:11:07 ok 17:11:21 How are folks doing on the dry runs? 17:11:26 #topic dry runs 17:12:11 lispflowmapping passed 17:12:16 i helped them 17:12:17 colindixon: OpenDOVE is finished with copyright/license headers. Affinity has begun to work on it 17:12:25 thanks phrobb 17:12:34 phrobb: I'll update the sheet 17:12:38 on controller we have some pending issues 17:12:46 GiovanniMeo, IIRC I think we passed the dry-run as well (as much as we could) 17:12:51 or is there more we can do? 17:12:56 we still need to remove the faulty licenced dependency 17:12:56 you can do more now 17:12:58 IGNORE_SNAPSHOTS ? 17:13:00 yes 17:13:02 GiovanniMeo: thank you :) 17:13:02 openflowjava passed dry-run 17:13:21 edwarnicke, GiovanniMeo: we have the following projects in attendance and they can report in. we can add it as a task even: projects in attendance: openflowplugin, openflowjava, opendove, yangtools, bgpcep, ovsdb, controller 17:13:28 per michal_rehak ofplugin dry run done 17:13:33 I have one last fix for controller dry run, it's queuing now 17:13:53 lim here for lispflowmapping 17:14:05 #info bgpcep has competed dryrun with ignored snapshots 17:14:23 #info opendove will revisit dry run later today 17:14:58 #info controller has dryrun running now (with fix for previously found issue) 17:15:07 #info ovsdb ignore_snapshot is being done. the rest are completed. 17:15:18 and people can fill this stuff in 17:15:24 I created a row for dry run 17:15:42 #action projects to fill in dryrun status on the per-project tab 17:15:46 ok 17:16:22 #topic testing with new openflowplugin 17:16:33 do we want to add a row to this status sheet for that 17:16:34 ed, any plans for in unstable race condition that causes our integration test to fail from time to time? 17:16:35 probably 17:16:49 #info identified the problems that ovsdb was having with new of13 plugin 17:17:07 #info it sounds like VTN hit some issues with OF13 plugin testing and have filed bugs/informed people 17:17:12 #info last night VTN project reported that of13 doesn't work w/ 1.0 mininet and doesn't work w/ VTN w/ 1.3 mininet 17:17:24 #info there is a OF version transition happening in the presence of a devstack/openstack environment and is not being handled by the plugin. 17:17:35 #indo am working with the openflowplugin folks on that. 17:17:39 i think we should have a of13 <-> 1.0 mininet test in integration as part of standard default testing 17:17:51 can anyone from of{plugin|java} comment on the other issues? 17:17:58 cdub: that sounds like a good idea 17:18:07 (assuming we expect of13 to actually manage 1.0 ;) 17:18:29 LuisGomez: thoughts/plans on that? 17:18:40 oh, no LuisGomez... 17:19:15 yeah 17:19:15 We have just finished triaging bugs on that 17:19:27 It is looking like a large number of bugs rooted in a small number of root causes 17:19:38 edwarnicke: "that" == of13 <-> 1.0 ? 17:19:44 We had some fixes go in overnight 17:19:49 And will have more going in today 17:19:53 #info edwarnicke says that the openflowplugin/openflowjava team have just finished triaging bugs and trying to find the root causes 17:20:16 hey RobDolin 17:20:19 ok 17:20:22 so we should expect the integration builds to get stabalized? 17:20:28 Hey Colin 17:20:30 cdub: 'that' being overall OFplugin bugs, including the OF10 etc stuff 17:20:42 Gmainzer: For lispflowmapping? 17:20:47 Gmainzer, edwarnicke: is this a topic we should cover here (integration build instability?) 17:20:54 edwarnicke: ok, cool. any idea if there's standard test in the works? 17:21:11 to validate, and protect against regression 17:21:15 colindixon: It depends on what Gmainzer means :) 17:21:22 yes, for lisp 17:21:47 Gmainzer: When I last checked with goldavberg about 2 hours ago he said all was well 17:21:48 shall we finish the discussion on the new openflowplugin testing ? 17:22:13 Madhu, cdub, edwarnicke: I'd love to 17:22:41 i cannot finish the OF13 testing if we don't get these issues resolved 17:22:46 is anyone here from integration team that's helping with this testing? 17:22:53 i heard there is a rare (1out of ~100) integration test runs that failed on that race (after there was a fix) 17:22:55 i have updated the devs from all the projects involved yesteday 17:23:01 the question from cdub to edwarnicke (and the ofplugin/java team) is "do we have a plan to put a test in place that goes to OF1.0 via the new plugin" 17:23:03 waiting on response. 17:23:27 cdub: I don't think so 17:23:47 colindixon: hmm...ok, we'll have to follow this up on the list i'm afraid 17:23:57 cdub: I believe LuisGomez is running the existing integation tests. We have been working to bugs from that for the last week or so 17:24:26 #action cdub to follow up on the list about getting a test set up to try to cover the more recent OF10 <=> OF13 bugs that have come up 17:24:30 We also have a bunch of folks (various groups, Ericcsson, Cisco, TSC, etc) running testing manually 17:24:40 colindixon: aye aye 17:25:00 #topic download page and artifacts 17:25:10 shague__ and phrobb, do we have a quick update here? 17:25:30 Gmainzer and edwarnicke, tell me if you we need an IT stability topic or not and I'll queue it up 17:25:53 wiki has the results for the download artifacts: https://wiki.opendaylight.org/view/Simultaneous_Release:Simultaneous_Release_Plan_2014:DownloadableReleaseArtifacts 17:26:15 #link https://wiki.opendaylight.org/view/Simultaneous_Release:Simultaneous_Release_Plan_2014:DownloadableReleaseArtifacts this wiki page has the results for the download artifacts from shague__ 17:26:21 Gmainzer: Do we need a topic? All open issues I have for lispflowmapping are closed... let me know if otherwise 17:26:22 Madhu: putting ofp-1.3 into integration requires porting of automatic tests to python 2.6 - that is currently not finished 17:26:45 I can send that to discuss if we want more eyes on the wiki 17:26:51 michal_rehak: Note: that's for RESTCONF. The existing NSF tests do not need porting to python 2.6 17:26:57 michal_rehak: is this for me ? 17:27:04 shague__: that's probably good 17:27:06 shague__: sent you a note about rpms for opendove -- pls have a look when you get a chance 17:27:07 phrobb: are you around? 17:27:10 edwarnicke: ok 17:27:12 michal_rehak: i don't remember asking that request 17:27:31 #action shague__ to send a mail to the discuss list about this to get more eyes on the wiki 17:27:40 ashaikh: yes, looking at it, get back to you soon 17:27:41 Madhu: sorry - was to colindixon 17:27:49 michal_rehak: sure. 17:27:53 No status on the download page mock up. Trying to get one now. 17:28:09 #info phrobb says "No status on the download page mock up. Trying to get one now." 17:28:21 ok that's not the most pressing issue for me right now, so that's fine :p 17:28:48 michal_rehak: ok, can you and cdub and maybe others follow up on what we can do to make this faster/better than manual testing a bug recovery 17:29:24 #topic additional testing 17:29:26 michal_rehak: let's connect after the meeting 17:29:46 cdub: ok 17:29:57 GiovanniMeo: edwarnicke we need to get the node-type/node-id issue fixed ASAP for AD-SAL / MD-SAL open flow convergence 17:30:02 Gmainzer: looks like a race between config-manager and netconf-server. I'll try to get someone on it, though ETA may be Monday morning CET :-/ 17:30:04 do we have an reports on canary testing and platform testing 17:30:06 Madhu: Its on my plate 17:30:12 that is a big ticket item for my testing the new plugin 17:30:15 thanks edwarnicke 17:30:17 ashaikh, RobDolin 17:30:18 I've got a fix about 3/4 of the way done 17:30:32 would be happy to try that if u want 17:30:34 Madhu: Don't thank me till its merged and verified working :) 17:30:50 thanks for keeping it on the list :) 17:30:52 Madhu: Much appreciated :) I will point you to the patch when I push it 17:31:02 colindixon: i have a note to RobDolin re canaries for windows -- otw will find some other way , maybe with fewer variants 17:31:38 #link https://wiki.opendaylight.org/view/Simultaneous_Release:Simultaneous_Release_Plan_2013/plaform_matrix this is the testing we wanted to get done 17:32:00 #action ashaikh to follow up with RobDolin to see if we can't get this slimmed down and figure out how to get this done 17:32:19 Are there other folks who may be able to help with Windows testing? 17:32:20 colindixon: i am going to try to do a SLES test today if possible 17:32:37 ashaikh: that sounds good, I'm just trying to make sure I corner all the issues I can here 17:32:59 #help RobDolin is looking for others to help with windows testing, colindixon looks at regXboi 17:33:16 no can do folks 17:33:22 regXboi: that's fair 17:33:45 #help also could use help in general even if it's just to help ashaikh prune the list to something reasonable and then run things 17:34:01 #topic release reviews 17:34:08 keep in mind that some of the tests are already being run by integration as a matter of course 17:34:32 #info just a friendly reminder that we will have release reviews next week and there's a mostly pro-forma wiki page each project will have to fill out 17:34:37 ashaikh: good point 17:34:49 #topic Documentation 17:35:25 RobDolin: did you want to go over anything here? or just follow up offline 17:35:32 Re: Documentation, I've added some links to: https://wiki.opendaylight.org/view/Release/Hydrogen 17:35:44 I've noticed that much of our documentation is per-project rather than per edition. 17:36:02 #link https://wiki.opendaylight.org/view/Release/Hydrogen RobDolin has added some links on this page 17:36:55 What do folks think about putting-up templtes per-project (rather than per-edition) and then encouraging contributors for each project to populate 17:37:12 ? 17:37:16 we should have a page that links the per-project into edition 17:37:25 so folks can find what's in the edition they've pulled 17:37:29 but otherwise I'm ok with that 17:37:58 ashaikh, regXboi, Madhu, rovarga, oflibMichal, michal_rehak, prasanna, edwarnicke, abhijitkumbhare, don't #info your response, but I just want to call attention to people representing projects that they will need to create a release review document by next week 17:38:19 is there a template? 17:38:26 RobDolin: if there's going to be per-project stuff included in each edition, we can create a subpage for it and then transclude it 17:38:34 I was going to wait until that became an issue 17:38:47 @ColinDixon, I think that might be a good approach. 17:38:59 colindixon: is there a template for the release review doc? 17:39:15 We can have per-project documentation and then include it in the relevant edition(s) 17:39:18 RobDolin: I was *really* hoping to at least get the core parts of the User Guide to be written in non-per-project way 17:39:22 like how to install 17:39:33 And how to run 17:39:35 regXboi: there is, here: https://wiki.opendaylight.org/view/Sample_Release_Review 17:39:51 @ColinDixon - I agree, but I'm worried we don't have folks stepping-up 17:39:54 And description of the features of the edition 17:40:01 #info RobDolin notes that much of our documentation is per-project, but our docs are oriented more per-edition 17:40:32 colindixon: thanks 17:40:43 I think that it becomes a matter of documenting the features of the edition, which are often per project... we may have some transclusion to do there 17:40:46 Could we identify volunteers to at least write the per-edition install guides? 17:40:56 #info colindixon thinks we need to have at least how to install and run with a mini task walkthrough per-edition (with the possible exception of virt which will likely need a mini task per virt tech, i.e., VTN, OVSDB, OpenDOVE) 17:41:01 RobDolin: I agree 17:41:03 RobDolin: I am all over that (and other doc things) after artifacts are cut 17:41:22 @EdWarnick - Awesome! 17:41:34 I'd also like to characterize the 'per project' sections as 'features' sections... rather than making us look like just a shambling mound of projects 17:41:35 #help do we have volunteers to help with writing the per-edition install guide? 17:42:21 mlemay contacted me privately yesterday to say that he and some of his people would try to get helping on the wiki with docs today 17:42:31 @ColinDixon - Maybe someone to volunteer to do the Base Edition install guide today? 17:42:40 anyone? 17:43:07 I can volunteer to take a stab, but not likely until tomorrow 17:43:42 #action colindixon to send a mail to discuss begging for somebody to volunteer to do the base edition install guide, but will likely take a stab himself in the worst case :-/ 17:44:10 RobDolin: That would be great, as I think other installation guides are just layered on top of that 17:44:21 #agreed WE NEED LOTS OF THE HELPS WITH THE DOCS!!!! 17:44:45 um 17:44:47 ok, are there any other topics people wanted to bering up 17:44:58 the virt edition install guide needs to be broken down by flavor 17:45:08 because it's wildly different between the types 17:45:22 regXboi: yes, we may just wind up having 3 sections there 17:45:32 we're going to need volunteers from each project to *really* help there 17:45:42 obviously we're writing the opendove install guide 17:46:24 #info regXboi correctly points out that the virtualization edition is going to likely have 3 different parts: one for each virtualization technology, we need to decide whether and how to break it up 17:46:48 regXboi: I see both the openDove and VTN extra pieces as part of the overall virtualization edition install guide (you need not agree with me... but those are my 2c) 17:46:49 I'm going to leave that for VTN, OVSDB, and OpenDOVE to work out since you guys will be producing that stuff 17:47:17 @ColinDixon - Count me in to buy you a drink for taking the first cut at the Base Edition Install Guide. 17:47:22 * edwarnicke seconds that ultimately the decisions on virtualization docs are in the hands of the guys doing the work 17:47:33 edwarnicke: I expect to find that the sections are independent 17:47:37 And so it begins... drinks4docs :) 17:47:39 +1 just that am not volunteering at this point. networkstatic ? 17:47:52 beyond the "download virt edition" part :) 17:48:30 sorry, in ofplugin room on another uissue, catching up 17:48:33 #agree the decision as to how to deal with breaking up the documentation for the virt edition should fall to it's stakeholders: OVSDB, VTN, and OpenDOVE 17:48:40 ok, any other topics 17:48:56 otherwise I move to get back to real work and sprinting toward 1/27 17:48:56 rogr, will bird dog docs for ovsdb 17:49:13 networkstatic: thanks bro 17:49:23 course sir 17:49:27 just 1 issue 17:49:30 #action networkstatic to "bird dog" docs for ovsdb 17:49:36 hehe 17:49:39 i ran the dry-run with IGNORE_SNAPSHOTS 17:49:44 ok? 17:49:46 hitting some issue with tagging 17:49:59 tykeal: or anyone can help ? 17:50:12 Madhu: am trying already! ;) 17:50:13 Madhu, GiovanniMeo_Away, tykeal, edwarnicke: is this worth chatting about for a few minutes here? 17:50:22 ok 17:50:37 it sounds like there are people on this 17:50:39 thanks tykeal 17:50:58 is anyone else having issues? any other topics before I hit the gong of meeting ending? 17:51:06 gong it 17:51:08 tykeal: Not an immediate complaint.. but looking at Jenkins... controller Jenkins may need more hardware going forward 17:51:15 Its working fine, but the queue is long 17:51:31 gonging 17:51:35 #endmeeting