09:00:13 #startmeeting blazar 09:00:14 Meeting started Tue Aug 22 09:00:13 2017 UTC and is due to finish in 60 minutes. The chair is hiro-kobayashi. Information about MeetBot at http://wiki.debian.org/MeetBot. 09:00:15 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 09:00:18 The meeting name has been set to 'blazar' 09:00:32 #chair priteau 09:00:33 Current chairs: hiro-kobayashi priteau 09:00:43 #topic RollCall 09:01:11 hi 09:01:23 o/ 09:01:30 hi all 09:01:39 o/ 09:01:44 o/ 09:01:44 hi masahito. You can attend only 10 mins today, right? 09:01:48 sorry, I need to go in 10 mins 09:01:57 Hi all 09:02:03 hi 09:02:09 #topic agenda 09:02:19 Todays agenda is 09:02:26 1. Pike release status 09:02:28 and I won't join next meeting. 09:02:34 2. Denver PTG 09:02:40 3. AOB 09:02:44 Anything else? 09:02:56 masahito: np 09:03:14 tc interview 09:03:44 ok 09:03:53 4 tc interview 09:04:10 masahito which one do you want to start? 09:04:57 1 , 4 then 2 if possible 09:05:02 Ok 09:05:13 #topic Pike release status 09:05:38 There are only 2 patch waiting for being merged. Great progress! 09:05:49 https://review.openstack.org/#/c/491744/ 09:05:54 Yes, great job everyone 09:05:56 https://review.openstack.org/#/c/482811/ 09:06:23 I have just rebased 482811 and I am planning to +2 09:06:32 I left comments for 491744 09:06:36 right, Pike is good release 09:06:53 priteau: thanks for your comment for 491744. I'll handle it after this meeting. 09:06:58 Thanks hiro-kobayashi 09:07:24 Anything to discuss here? 09:07:31 > all 09:07:55 I'll cut stable/pike branch once both patches are merged 09:08:09 masahito: thanks! 09:08:58 Deadline will be Aug. 30, right? 09:09:02 Great 09:09:16 right 09:09:36 OK. 09:09:46 Anything else? 09:10:03 If nothing, let's move on 4 09:10:20 #topic TC interview 09:10:53 masahito: Can you introduce this topic? 09:11:22 TC team offered this team a final interview in next PTG 09:11:49 to official project. 09:12:03 Great! 09:12:12 thanks, date/time fixed yet? 09:12:16 the day is Mon. or Tue. 09:12:16 great 09:12:40 Is that a face to face interview? 09:12:55 I'll be there from Tue. so it would be Tue. 09:13:06 priteau: yes 09:13:30 Is there something to prepare? 09:14:00 nothing are required 09:14:03 can we join or is it you and the TC? 09:14:38 if you are in there, it's ok 09:14:51 they ask 09:15:31 us what we need if we are not official by the time. 09:16:20 if we're accepted, they tell us what we need to do next. 09:16:35 great, thanks 09:16:53 thanks 09:17:33 no -1 on gerrit so far :-) 09:18:02 #link https://review.openstack.org/#/c/482860/ 09:18:29 of course, i have to prepare answers of faq for the meeting :-) 09:18:58 that's from my side 09:19:17 masahito: thanks! Let us know if any help is needed 09:19:34 Anything else about this topic? > all 09:19:35 got it 09:20:24 #topic Denver PTG 09:20:49 https://etherpad.openstack.org/p/blazar-ptg-queens 09:21:09 masahito has arranged time slots for our meeting. thanks masahito! 09:21:39 Please see L4-14 09:22:21 1st meeting is for all of us. 2nd is for attendances or European time zone. 09:22:55 OK. maybe I will sleep at that time ;-) 09:22:56 I suppose we won't have an IRC meeting at 0900 UTC on the 12th? 09:23:04 hiro-kobayashi: sorry, the midnight meetings... 09:23:09 masahito: np 09:23:37 priteau: yes 09:24:04 uh? same time? 09:24:57 masahito: Since we have the PTG meeting, the regular IRC meeting is not necessary and will be difficult to attend from Denver anyway? 09:25:17 yes. so 09:25:36 I planned to cancel it 09:25:48 OK 09:26:54 Please recheck your attendance at L15-24 > all 09:26:58 official teams have a priority for the space. 09:27:38 the space could be change. 09:28:15 OK 09:28:50 but if the time slots are okay to you I share it to ML next week 09:30:47 They are good for me 09:31:43 thank all! I need to go. I'll read the meeting log later. 09:31:55 masahito: bye! 09:32:04 I'm sorry I was disconnected 09:32:04 bye 09:33:57 If you have anything to discuss at PTG, please right it down etherpad 09:34:12 anything else? 09:34:38 #topic AOB 09:35:02 Oh, I'm not a chair now. priteau: could you change the topic? 09:35:09 #topic AOB 09:35:11 #chair hiro-kobayashi_ 09:35:11 Current chairs: hiro-kobayashi hiro-kobayashi_ priteau 09:35:21 priteau: thanks! 09:35:31 I am requesting comments for this patch: https://review.openstack.org/#/c/494476/ 09:35:57 It adds state information to leases which is displayed by the CLI and Horizon dashboard 09:36:11 Sorry, I couldn't check it yet 09:36:13 It's not 100% complete but I would like to have your opinion before I spend more time on it 09:37:46 The status of lease should be consistent with the status of reservation 09:37:50 IMHO it is important to keep the status and status reason updated. 09:38:30 without having looked at the patch in detail yet I feel it is going in a valid direction 09:38:47 hiro-kobayashi_: the current DB format is different between leases and reservations 09:39:01 For reservation, it's just one "status" field 09:39:32 priteau: yes. Shouldn't them be aligned? 09:39:46 For leases, there is an "action" field and a "status" field, which according to the existing code was meant to be used that way 09:40:16 I would agree to make them more similar, though that will likely require DB schema changes 09:40:37 priteau: have you looked at v2 api? what is your intention here? 09:40:48 #link http://blazar.readthedocs.io/en/latest/restapi/rest_api_v2.html 09:41:57 bertys: I didn't notice it was documented in v2, but yes my patch pretty much matches that 09:42:46 bertys: thanks for your info! 09:42:55 np 09:43:18 The main question is, do we want any kind of lease state info in Pike, while we develop an alternative state model for Q? 09:44:25 I think we don't need it in Pike 09:44:49 Gut it's very important task for Q 09:44:59 s/Gut/But 09:45:57 As we discussed at Boston, we should clearly define state machine. 09:46:06 hiro-kobayashi: +1 09:46:34 Yes, it 09:46:46 it is important to work on the state machine 09:47:12 in Pike we could already start working on the mechanisms. if we change state model for Q, for this work do you expect major changes to the code or minor edits only? 09:48:12 GeraldK: Pike is going to be branched very soon (maybe later today). We can backport bug fixes, but it's preferable to avoid backporting major changes. 09:49:46 priteau: agree. so you expect your patch and other work on lease state info NOT to be included for Pike? 09:50:48 priteau: maybe I misunderstood your above comment where you asked whether to have state info in Pike or not 09:50:52 My patch doesn't change any of the schema and intends to use the existing mechanisms, so you can almost consider it a bug fix ;-) But I would include it in Pike only if everyone else was OK with it 09:52:12 I'll check the patch in detail and comment on Gerrit. Basically I agree with your approarch, priteau 09:52:12 If the rest of the community prefers to state info for Pike and instead focus on the new mechanism for Queens, we can do that. 09:53:17 the parts where you update the state based on events seems like something that is independent from the state model and would be kept even if we update the state model, right? 09:54:15 GeraldK: There would still be code in the same place, but the code would do something different I guess. 09:54:36 e.g. instead of setting action=START and status=COMPLETE, it would set status=started 09:54:50 Or status=active 09:55:07 I think we need a blueprint for the new mode l:-) 09:56:08 #action review https://review.openstack.org/#/c/494476/ and discuss lease states 09:56:10 priteau: okay I see. a lot of rework would be needed. 09:56:17 yep 09:56:44 And there are some edge cases to consider, e.g. if a lease has multiple reservations, some start correctly and one doesn't 09:56:50 What's the lease status in that case? 09:57:30 error or something? 09:57:44 A lot of works to do :-) 09:57:55 priteau: Anyway, thanks for starting this work! 09:58:02 np 09:58:22 Let's keep discussing on Gerrit 09:58:42 Anything else? a few minutes left ... 09:59:27 OK. Let's end today's meeting. 09:59:30 thanks all! 09:59:33 Thanks everyone 09:59:35 Good bye. 09:59:43 Good bye! 09:59:46 bye 09:59:54 #endmeeting