14:32:14 #startmeeting weekly Lithium IRC sync 14:32:14 Meeting started Wed Mar 18 14:32:14 2015 UTC. The chair is colindixon. Information about MeetBot at http://ci.openstack.org/meetbot.html. 14:32:14 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:32:14 The meeting name has been set to 'weekly_lithium_irc_sync' 14:32:16 * alagalah_ looks around... 14:32:21 #info alagalah GBP 14:32:26 #undo 14:32:26 Removing item from minutes: 14:32:27 #topic agenda bashing and roll call 14:32:28 * regXboi notes oh there *IS* a meeting :) 14:32:38 #info regXboi NN 14:32:43 #info Prem VPNService 14:32:45 #chair regXboi alagalah_ phrobb- 14:32:45 Current chairs: alagalah_ colindixon phrobb- regXboi 14:32:56 #info colindixon for TTP, docs, and TSC 14:33:09 #link https://wiki.opendaylight.org/view/Simultaneous_Release:Lithium_Release_Plan#Cross_Project_Meetings agenda in it’s usual place 14:33:10 #info abhijitkumbhare OpenFlow plugin (can stay only some time) 14:33:12 * regXboi notes we should have named it neutron and northbound just so he could info in as NaN :) 14:33:12 #info gzhao release 14:33:52 #info fabiel for Persistence Store Service 14:34:07 #infi LusiGomez for integration 14:34:19 #info LuisGomez for integration 14:34:44 anyone have topics they want to raise? 14:35:25 #topic upcoming deadlines 14:36:10 #info M3 for offset 2, M4 for offset 0 due tomorrow 14:36:23 #info that’s API freeze for offset 0 14:36:35 #info and functionality freeze for offset 2 14:37:01 #info Helium 3 test and bug verification should be complete today 14:37:22 #link https://wiki.opendaylight.org/view/Tech_Work_Stream:Main#Information_From_Past_Meetings the March 16 TWS meeting has info on how to get your system tests started (required for M3) 14:37:54 #link https://wiki.opendaylight.org/view/CrossProject:Documentation_Group:Lithium_Project_Documentation_Requirements#Step-by-step_Guide this page has information on how to submit your documentation outliens (required for M3) 14:38:11 #chair gzhao 14:38:11 Current chairs: alagalah_ colindixon gzhao phrobb- regXboi 14:38:29 #topic questtions, comments, concerns, announcements, etc. 14:38:34 * tbachman wonders if colindixon undid alagalah_ ’s #info above 14:38:38 * colindixon opens up the floor 14:38:45 (for info-ing in) 14:38:53 yes - on purpose 14:39:09 alagalah_: you should re-info 14:39:12 regXboi: ack 14:39:18 #info alagalah_ for GBP (I undid it earlier because it was before the topic, this is redoing it) 14:39:18 #info alagalah GBP 14:39:21 sorry 14:39:26 just leave it, not that much dirt 14:39:28 colindixon: no worries! 14:39:43 My apologies, apparently my kid thought it would be acceptable to go to school without brushing her teeth (should have let her) 14:39:46 so, does anyone have any questions, concerns, annaouncements or anything? 14:39:53 sorry 14:39:58 timezone change 14:39:58 * regXboi listens for chirp chirp chirp 14:40:06 * tbachman smells stinky breath 14:40:07 colindixon: quick question 14:40:08 colindixon: Q 14:40:11 ttkacik1: you first 14:40:46 colindixon: does removal of deprecated APIs (from previous release) need to happen before API freeze? or do we need API Freeze Waiver? 14:40:54 I have a question about superpom, do we reach a decision with rovarga 14:41:29 ttkacik1: I would think it should happen before API freeze, but I can be argued around to the other position 14:41:40 ttkacik1: I think I would tend to agree with regXboi 14:42:00 ttkacik1: any other projects depend on those API? 14:42:04 #info ttkacik1 asks if removing APIs that were deprecated in a previous release needs to happen before API freeze 14:42:16 #info colindixon and regXboi say they think that probably makes sense 14:42:22 still TTP left out for composite nodes 14:42:44 ttkacik1: I thought you were going to submit a patch? 14:42:52 you have it there 14:43:11 was bit tricky to figure it out 14:43:14 colindixon: I have my question after ttkacik1 .. want me to queue it up for addressing when you are done? Also looks like gzhao is in the Question Q (QQ) :) 14:43:55 problem is that some of other projects still sports uses other deprecated APIs 14:44:14 but I am fine with API Freeze Waiwer for removal 14:44:49 (just want to get Hydrogen code out from Lithium - since that backwards compatibility is entry barrier for people going to yangtools or controller) 14:46:49 #action colindixon to review the patch to TTP to migrate away from composite nodes 14:46:54 alagalah_: take your topic away 14:47:02 colindixon: Ach 14:47:22 colindixon: So are we doing a IRC meeting "scratch and sniff" on offset 2 M3 or is it just email ? 14:47:43 ttkacik1: is there a mail to the projects still using the APIs you’d like to remove? 14:48:17 * Prem wondering if the QQ is full or he can get his Qn in 14:48:20 * alagalah_ recalls the IRC review from He but can't recall what gate it was 14:48:28 alagalah_: I’m not sure I follow 14:48:48 Prem: please add questions to the queue 14:49:03 alagalah_: I think it’s just an e-mail 14:49:06 colindixon: In He, projects had a Mx gate review with the TSC over IRC. We had timeslots etc assigned.. I can't recall if that was for a later gate or if we are doing away with it 14:49:20 alagalah_: that was for M5/RC0 I think 14:49:42 colindixon: Ack... then our email is just about queued up ready to go... I ceded the floo 14:49:43 the composite node ones - they have already migrated / there are patches 14:49:43 florr 14:49:49 floor sigh... long day already 14:50:04 for other deprecated APIs (odl data store APIs) still need to figure out if there are such projects 14:50:24 #info alagalah_ asks if there is going to be a TSC review or just an e-mail for M3 14:50:57 #info colindixon says that there will just be an e-mail the only time there’s a review with slots is for the release review between M5 and release (where TSC attendence is optional) 14:51:01 Prem: did you have a question 14:51:25 colindixon: Yes, there are certain items that are in work in progress 14:51:47 How far the ETA can be :) 14:52:26 gzhao: ? 14:52:45 Prem: I can say we went +1 week on M3 for NN as an offset 1 and nobody complained 14:52:46 Prem: for now, I think the idea is report the honest truth about where you are and what you could do to get back on schedule 14:52:53 well ETA to 3/31 is a little to long 14:53:05 but ETA is ETA to complete 14:53:32 Prem: from there the TSC and release staff will figure out what risk there is and decide how to try to help or not 14:53:38 Prem: of course, we did flag ourselves as yellow when we did it 14:53:57 Prem: does VPN service hae any one depending on it? 14:54:23 colindixon: Nope. As of now we don't have any depedency on us 14:55:15 #info Prem asks how far out a deadline can be pushed with an ETA and have it be “reasonable" 14:55:44 #info gzhao and colindixon say to report where the honest progress is and how far away things are froming being done and the TSC and release staff will try to figure out if there’s a problem 14:56:25 colindixon: To make the VPNservice solution truly end to end, working with Openstack and GBP and hence wanted to some help 14:57:21 Prem: the short version (in my opinion) is this: if your project isn’t blocking others and will actually deliver something you feel is useful in the release in a reasonable time-frame, it’s probably fine 14:57:58 if you start to block others either directly (the depend on you) or indirectly (because you’ve broken the release build), then things get a lot harder 14:58:04 colindixon: Thanks and that helps! 14:58:07 but it doesn’t sound like that’s the case 14:58:28 Prem: also, we’d really like to see an understanding of how you will get across the finish line 14:59:30 Prem: so, will you deliver less than you expected? or are you getting extra help? or what? 14:59:34 if that makes sense 14:59:50 not to cut Prem off, but anything else people wanted to bring up? 14:59:52 colindixon: We are good to go as a standalone module with needed APIs defined than can be used to define 15:00:16 * regXboi takes off NN hat and puts on TSC hat 15:00:54 ttkacik1: I would say that if you send a notice that you will seek an API freeze waiver about removing deprecated APIs and list the affected projects, that would be the right thing 15:00:58 Prem: can you explain your "end to end" statement above? 15:01:47 #chair tbachman 15:01:47 Current chairs: alagalah_ colindixon gzhao phrobb- regXboi tbachman 15:01:59 * tbachman pops in 15:02:12 Sure. VPNService can be configured with needed parameters through API. What I meant end to end was - expecting these parameters to flow in from Openstack 15:02:16 * tbachman reads scrollbak 15:03:06 Prem: I'm going to push you a bit harder on that statement - how are you getting those parameters via openstack? 15:03:25 colidixon: I'm heading for a particular place here 15:03:39 colindixon: so bear with me for a few more minutes 15:03:51 #info regXboi|tsc_hat asks if Prem can explain what he means by making the VPNservice solution truly end-to-end 15:04:07 As of today, we don't have parameters other than L2 coming in from Openstack 15:04:27 #info Prem says VPNService can be configured with needed parameters through an API. By end-to-end — he was expecting those parameters to flow in from OpenStack 15:04:37 Prem: that's the what - I'm asking the how 15:04:49 #info regXboi|tsc_hat asks how they are getting those parameters via openstack 15:05:14 #info Prem says they only have L2 parameters coming from OpenStack 15:05:25 #info regXboi|tsc_hat asks how they come in from OpenStack 15:05:36 regXboi|tsc_hat: Prem I don't want to speak for Prem, but he and I are sitting down tomorrow to work out if OS GBP and ODL GBP together can get him over the line. 15:05:51 alagalah_: that's what I'm trying to call out 15:06:02 regXboi|tsc_hat: Prem Or some combination there of etc etc etc ... fullness of time, all things considered... 15:06:38 because in the spirit of full disclosure we have to point out to our customers that if VPN doesn't use NN, they will have to use a different part of OpenStack... 15:06:52 * regXboi|tsc_hat takes off tsc_hat now 15:07:31 * gzhao curious does vpn services get from openstack through NaN? 15:07:57 #info alagalah_ says that he and Prem are sitting down tomorrow to work out if OS GBP and ODL GBP together can get him over the line. 15:07:58 gzhao: Yes, that is how we started 15:08:10 Prem: thanks 15:08:10 #info regXboi says in the spirit of full disclosure we have to point out to our customers that if VPN doesn't use NN, they will have to use a different part of OpenStack... 15:08:46 all: tbachman's info is where I was trying to go - I'm done 15:08:59 so, I think it’s pretty clear that we want to track what’s happenign with VPN service, just to make sure they can have the best chance of getting the most out of the release 15:09:03 * tbachman notes that regXboi’s hat suffix is off 15:09:21 #info gzhao asks Prem if VPN services gets from openstack through NaN 15:09:21 thanks for letitng us know Prem and it sounds like you’re doing the right things 15:09:27 #info Prem says yes, that’s how they started 15:09:31 ok 15:09:35 so, any more topics? 15:09:43 Prem: I can see the dependency diagram changes for vpn service. 15:09:48 #info colindixon says we want to track what’s happenign with VPN service, just to make sure they can have the best chance of getting the most out of the release 15:10:05 #info gzhao says the dependency diagrams need to change for VPN service 15:10:21 gzhao: Yes 15:11:01 ok 15:11:07 last call for topics 15:11:09 going once 15:11:27 colindixon: you know I have one :) 15:11:35 going twice 15:11:57 #topic muffins (too early for cookies) 15:12:09 * tbachman grabs muffin, before meeting ends 15:12:19 gone 15:12:22 #endmeeting