Monday, 2016-12-12

*** zhurong has joined #cloudkitty01:18
*** liujiong has joined #cloudkitty01:41
*** sacharya has joined #cloudkitty02:12
*** sacharya has quit IRC02:24
openstackgerritchenyingnan proposed openstack/cloudkitty: Replace six.iteritems(iter) with iter.items()  https://review.openstack.org/40959003:32
*** sacharya has joined #cloudkitty04:24
*** sacharya has quit IRC04:29
*** sacharya has joined #cloudkitty05:26
*** sacharya has quit IRC05:30
openstackgerritchenyingnan proposed openstack/cloudkitty: Replace six.iteritems(iter) with iter.items()  https://review.openstack.org/40959005:49
*** bumblebee has quit IRC06:12
*** openstackgerrit_ has joined #cloudkitty06:23
*** openstackgerrit_ has quit IRC06:25
*** openstackgerrit_ has joined #cloudkitty06:26
*** openstackgerrit_ has quit IRC06:27
*** openstackgerrit_ has joined #cloudkitty06:34
*** openstackgerrit_ has quit IRC06:35
openstackgerritchenyingnan proposed openstack/cloudkitty: Replace six.iteritems(iter) with iter.items()  https://review.openstack.org/40959006:56
*** openstackgerrit_ has joined #cloudkitty06:59
*** openstackgerrit_ has quit IRC07:00
*** sacharya has joined #cloudkitty07:27
*** sacharya has quit IRC07:31
*** liujiong has quit IRC08:06
*** liujiong has joined #cloudkitty08:07
*** peschk_l has joined #cloudkitty08:28
*** peschk_l has quit IRC08:34
*** peschk_l has joined #cloudkitty08:45
*** zhangguoqing has joined #cloudkitty09:40
*** gpocentek has joined #cloudkitty10:00
*** zhurong has quit IRC10:02
*** liujiong has quit IRC10:18
*** pabardina has joined #cloudkitty10:58
*** zhurong has joined #cloudkitty11:54
*** zhurong has quit IRC12:06
*** zhurong has joined #cloudkitty12:07
*** zhurong has quit IRC12:19
*** liujiong has joined #cloudkitty12:50
*** zhurong has joined #cloudkitty12:53
*** ivanromanko has joined #cloudkitty12:57
zhangguoqinghi13:03
zhangguoqingMonday 12th at 13:00 UTC13:04
*** jwcroppe has quit IRC13:06
sheeprineHi zhangguoqing13:07
zhangguoqingmeeting time now. :)13:08
sheeprineyup13:08
liujiong\o13:09
zhangguoqingHi liujiong. :)13:09
liujionghi13:09
*** zhurong has quit IRC13:11
*** zhurong has joined #cloudkitty13:12
sheeprinewhat time is it for you guys?13:13
liujiong9:00PM13:13
huatsHello !13:13
zhangguoqingsheeprine: 9:00PM13:13
huatsHow is everyone going ? ready to start ?13:13
liujionghi13:13
liujiong\o13:13
zhangguoqinghi huats, OK13:14
sheeprineDo we have any agenda for today?13:15
gpocentekhello13:15
zhangguoqingIn email: We'll be discussing the state of the various developments we are doing for Ocata.13:16
huatsExactly13:16
huatsI think it is important to do a small wrapup of the various blueprints and devs that are curently ongoing13:16
huatsBut also some ideas / improvment that will be done soon13:17
huatsOk so let's start then !13:17
huatsFirst of all13:17
huatsThank you all for being here13:17
huatsI know it is always complicated to find a good timeframe due the the various timezone....13:18
aolwashi13:18
peschk_lhello13:18
huatsI was really looking to have that kind of meeting13:18
huatsThe main reason is so that everyone have in mind one big idea : The ocata cycle will be really short13:18
huatsAs you might know ocata cycle will only last 4 months13:19
huatshttps://releases.openstack.org/ocata/schedule.html13:19
liujiongexactly13:19
huatsso it means the client and API are due (well at least freeze) on the 26th of january...13:19
huatsThat is why we need to have that kind of meetings...13:20
huatsIn order to have a better development and review experience13:20
sheeprineAnd we've just entered milestone 213:20
huatsand I mean for the whole community13:21
zhangguoqingyep, it's urgency.13:21
huatsno urgency. just to know it13:21
sheeprineI saw somes patches regarding reno and others modifying API endpoints13:21
sheeprineDo we have referenced this somewhere?13:21
huatssheeprine: let me finish please, we'll have some time to discuss that13:22
huatsWe have noticed in the past few weeks many reviews that might have been eased with a better workflow / testing environment13:22
huatsSo I think that we should start to push the experimental test we have to non voting (for a start). It will be a single devstack installation but it would help review...13:23
huatsThen Luka will work on the intergation of many 'internal' tests that we have @Objectif Libre to the one in devstack. So taht patches will be more tested before a need for a review13:24
huats(to optimize the time of reviers), allowing developpers to get a feedback (from jenkins) in a better time...13:24
liujiongthat would be great13:25
huatssorry when I mean Luka, I meant peschk_l13:25
huatswhat is your opinion on that ?13:25
* gpocentek agrees13:26
*** zhurong has quit IRC13:26
sheeprineJust to be sure I understand, you want to add a new job that deploys a devstack, but non-voting to avoid problems?13:26
huatscurrently it is on expermiental13:27
sheeprineIs the run failing ATM with the current codebase ?13:27
huatsso it is just to push it non-voting as a step13:27
huatsno it just runs perfectly13:27
*** zhurong has joined #cloudkitty13:28
huatsjust to do that step by step13:28
sheeprineDo we need to keep it as non-voting?13:28
huatswe can clearly discuss that too :)13:28
huatsI am not opposed of putting that voting13:28
sheeprineIf it's not failing with the current code base, I'll prefer it to be voting13:28
peschk_lsheeprine, maybe going experimental -> gate is not very OpenStack-compliant13:29
sheeprineso that we're sure contributed code is working13:29
*** jwcroppe has joined #cloudkitty13:29
sheeprinepeschk_l: We do a lot of things not openstack compliant ;)13:29
huatsany other opinions ?13:29
sheeprineThis one is not a huge deal I think13:29
peschk_lok, then it's ok for me if we make it voting13:30
sheeprineWe can then have more integration tests and run them against new code13:30
gpocenteklets try with a dummy review13:30
gpocentekif it works OK we can set it voting13:30
huatsOK for me13:31
sheeprineShouldn't experimental and non-voting do the same thing?13:32
huatsthey are separated AFAIK13:32
sheeprineI thought it was just a naming convention13:32
huatsI think it is not a big deal since we just decided to go for voting :)13:33
sheeprineYup13:33
huatsLet's move to the next point I wanted to address which is more "a preambule" :)13:33
huatsthe release number13:33
sheeprineJust push it voting and forget it13:34
huatsSo far cloudkitty is one of the only project on the Big Tent umbrella that still ha 0.XX number13:34
huatsI would be in favor of jumping to the regular number (like othe project). So it means ocata will be 5.013:34
huatswhat do you think of it ?13:34
huatsI really think it would be clearer and it gives a better message13:35
sheeprineYou know what I think about this ;) But we're not going to discuss this here again, you're the boss now.13:35
huats(especially in comparison to other projet)13:35
huatsnope. I want to have that kind of discussion open13:35
huatsopenly with everyone13:35
huatsI have my opinion for sure, but I don't want to force things13:36
sheeprine1.0 was supposed to have a defined number of features that are not yet in the code.13:36
liujiongI'm OK with that13:36
sheeprineMostly because 1.0 would mean that we publish the API as "stable"13:36
sheeprineand if these features are introduced after we'll need to modify the API deeply13:37
sheeprinemeaning a new version13:37
gpocentekwe never announced the features13:37
gpocentekand it seems more logical to follow the same versionning as other projects13:37
sheeprineIt was always anounced in the roadmaps of the previous versions13:37
sheeprinebut always postponed13:37
gpocentekI don't think that versions actually mean something in OpenStack13:37
sheeprineand we discussed this internaly too13:37
sheeprineJust shoot a random number then13:38
huatsI think that the important thing here is that the API can evolve13:39
sheeprineYep but if it's stable it should not13:40
huatsif it is not broken we can keep on the v1 of the API13:40
sheeprineYou can version it but you are creating new problems13:40
huatsI think it is important to differentiate versions numbers of the API and the software13:40
huatsthey can be related but it is not mandatory13:41
huatsso that is why I think we should follow the Openstack release number (well lilke other project)13:42
sheeprineI don't mean v1 = API v113:42
gpocentekI really think we should follow the openstack model, it's just more logical for end users13:42
sheeprinethen it'll just be major number = openstack version13:43
sheeprinewhich is meaningless13:43
gpocentekthe version is meaningless anyway13:43
aolwasIf CK releases are the same as Openstack, we may have numbering reflecting this13:44
huatsIs it ok that we decide to go as a release number since we entered the big tent ?13:45
sheeprineyeah but for example some of our releases didn't really bring new stuff on the table13:45
sheeprinehaving a new major number for this is nonsense13:45
sheeprinebut if it's the openstack way then go ahead13:45
huatsOK then. Let's go on that.13:46
huatsFinally before jumping to the current blueprints/patchs13:46
aolwassheeprine: a lot of compagnies now use the year as release number: 2015-1, 2015-2, 2016-1, etc ...13:47
huatsmany people asked in the past for a model of a patch (and commit message). We have been discussiing that with aolwas and we thought it would be great to have of kind of model. I think that  https://review.openstack.org/#/c/406180/13:47
huatscan be that...13:47
sheeprineaolwas: It was like that for openstack before but go dumped13:47
sheeprineaolwas: go figure13:47
huatsOk let's have a look with blueprints13:48
huatsDue to the very short cycle I think it is important to be able to focus on some features13:49
huatsSome blueprints have been marked as High in the mprotance are the one which are expected to land in Ocata13:49
huatshttps://blueprints.launchpad.net/cloudkitty13:50
huatssome have been superseed by other13:50
huatsI am not in favor of going through one by one13:50
huatsBut I'll be really interested to13:51
huatsgo on the High one13:51
huatss13:51
huats"get-total-price-by-tenant-and-user"13:51
huatsliujiong: you provided quite a lot of work on it, doing many exchanegs with peschk_l.13:51
huatsdo you have any blocker now ?13:52
liujiongright13:52
liujiongwe have some discussion on that. But recently, I have not much time implementing that.13:53
huatsThat happen, don't worry. Do you think you'll be able to continue on that soon ?13:53
*** zhurong has quit IRC13:53
liujiongyeah13:54
liujiongI think so.13:54
sheeprineDo we have patches conflicting with this one?13:54
sheeprineAs it might require quite some refactoring if it's postponed too late13:54
sheeprineAnd it's a nightmare ;)13:54
sheeprineIs this far from finished?13:55
huatsIt is quite a big patch, so there are some conflicts yes13:56
liujiongI remember dinghh has a patch related to that.13:56
huatsthat is why it would be great to merge it soon13:56
huatsliujiong: indeed, he will be the next one that we speak of :)13:56
liujiongOK13:57
zhangguoqinghttps://review.openstack.org/#/c/349779/ and https://review.openstack.org/#/c/393021/13:57
sheeprinePlus, when you wait too long you need to rebase and merge and with jenkins and all it's easy to have regressions (we've been throught them, painfull)13:57
huatsliujiong: do you think you'll be able to finish it let's say in the 2 next weeks ?13:58
huatsbecause with that short cycle it would be really really great...13:58
liujiong2 weeks should be too short for me.13:58
sheeprinehuats: Full freeze is end of Jan 2017, right?13:58
huats26/0113:58
liujiongOK, I'm fine if any of you want to take over that BP. Cause I'm really busy with other stuff recently.13:59
huatsliujiong: would you accept some help ?13:59
huatsI was about to propose you that13:59
liujiongYep, I'm fine with that.14:00
huatsthanks !14:00
peschk_lliujiong has already done a lot work on this, aolwas and I can look into it14:00
huatsmay be peschk_l can work a bit on that since he knows it a bit now14:00
huatsGreat !14:00
aolwasOK for me14:01
liujiongthanks, peschk_l, aolwas.14:01
huatsSo the next BP is related like we said : price-groupby-fields14:01
huatsdinghh: has been working on that14:01
huatsare you around ?14:01
zhangguoqinghuats: You may not known that the last month always busy in China. So, liujiong is very greate.14:01
liujiongthanks, guoqing. you really mean it.14:02
huatswe are greatfull for your hard work !14:03
huatsAaron, please wait that the liujiong to be merged before working on that again14:03
huatsit would be better :)14:03
huatsplease ping me if you have some questions...14:03
huatsSo the next BP is rating-volume-snapshot14:04
huatszhangguoqing: any blocker ?14:04
zhangguoqingdinghh may not here meeting now.14:04
zhangguoqingyep, I thing this BP has been OK, just for review.14:04
huats(zhangguoqing: he is connected so he'll be able to see thoses messages)14:04
zhangguoqinghttps://review.openstack.org/#/c/399831/14:06
huatsit should be reviewed quite easily14:06
huatsand soon14:06
zhangguoqingOK, and now, it's only for ceilometer collector, because the gnocchi has not metring volume.snapshot.14:07
huatswe (@objectiflibre) are having a sprint currently on CK. So it will be done during that14:07
liujiongguoqing has done great job in this cycle.14:07
zhangguoqingI will do that in gnocchi.14:08
huatszhangguoqing: great !14:08
aolwasyou need patches in gnocchi for that, right ?14:08
zhangguoqingaolwas: yes.14:09
huatsthis is something that we need to address : improvment of the gnocchi collector14:09
zhangguoqingIt will take more time to do it, a little complex.14:09
aolwasok14:09
zhangguoqinghuats: yes, it's very important.14:10
zhangguoqinghttps://review.openstack.org/#/c/408518/14:11
huatszhangguoqing: don't you think we should have a joint meeting with the gnocchi guys for that ? we know them very very well...14:11
huatsdo you think it would help ?14:11
zhangguoqingNo needed now.14:12
huatsok14:12
huatsplease let me know it is teh case14:12
huatsNext BP is "user-experience-improvement"14:13
huatsaolwas: do you have any blocker ? any help needed ?14:13
aolwasfor API and client, I'm ok14:13
aolwaswe just have to look how we improve the horizon dashbard14:14
huatsI know pabardina has planned to help you for horizon14:14
huatsso it would be OK14:14
aolwasgreat14:14
zhangguoqingaolwas: Good work.14:14
aolwasthanks !14:14
huatsNext BP is "maintain-releasenote"14:15
huatsit is something that has been tackle for every OpenStack projects14:15
zhangguoqinghttps://review.openstack.org/#/c/389669/14:16
huatsany opinion ? ideas ? deepack has not finished it. i would be in favor of contacting him and see with him if he has any issue or if want some help (I would do it then)14:17
huatsand opinion ?14:17
sheeprineWe need them14:17
zhangguoqingIt is just only deal with the releases notes.14:17
zhangguoqingfor netwon14:18
sheeprineAnd with the user experience modifications the API will be modified so we need to at least track this14:18
liujiongagree14:18
zhangguoqingyes, should be merged before the 'user experience'.14:19
huatsyep14:19
huatsSo I will contact him right now14:19
huats(after the meeting) and make sure that we get merged soon14:20
huatsI will finish it myself if needed14:20
huatslast BP for ocata "rating-for-second-level"14:21
zhangguoqingGreat, after that I will finished the python-cloudkittyclient and cloudkitty-dashboard. :)14:21
aolwaszhangguoqing: is your patch ready for review ? it's still marked as WIP14:22
zhangguoqingThe core code of "rating-for-second-level" has been finished.14:22
zhangguoqingbut the new ceilomieter change, so it's should be updated and tested.14:22
zhangguoqingIt will take more time, and hard to review.14:23
aolwasdo you think it's reasonable to look at it for ocata ?14:23
aolwasor should we postponed it to pike ?14:24
zhangguoqingNot sure.14:24
huatszhangguoqing: when do you think you'll be able to finish it ?14:25
huatsdo you have an easy way to test it ?14:25
huats(like peschk_l did with liujiong with a patch we have mentionned earlier)14:25
zhangguoqingI have wrote the unit tests.14:26
aolwasI'll try to look at it next week and provides you early returns14:27
zhangguoqingThe test maybe easy, but getting events from ceilometer has some problems.14:28
aolwaswe'll see then how we can go on for ocata integration (or not)14:28
aolwasproblems on ceilo side or CK ?14:28
huatsI think we should take a decision let's say first week of january if we want to push it or not. What do you think ?14:28
zhangguoqingproblems on ceilometer.14:29
huatsok14:30
huatszhangguoqing: and aolwas I tihink you should continue that after together14:30
huatsso we have finished the blueprints that are noted Hig14:30
huatsh14:30
huatsany other topic you want to discuss to include during ocata ?14:31
sheeprineJust as a quick reminder, for those of you who wants to contribute new blueprints. There is a new repository for this cloudkitty-specs. It's how it should be done as it helps people review and contribute to them, the openstack way.14:31
liujiongright14:31
sheeprineI hope we'll move on to using this in the future14:31
huatssheeprine: I agree14:31
aolwasme too14:31
zhangguoqingfirst week of january, I think it's hard to me. I'm busy on thesis proposal for my master degree.14:32
liujiongI have a patch initializing the repo. https://review.openstack.org/#/c/402310/14:32
huatsI really think it was too complicated to move on during ocata with the 4 months duration14:32
*** ivanromanko has quit IRC14:32
liujiongdidn't get any response.14:32
huatsbut YES we should move on there ...14:32
sheeprineliujiong: I've been busy lately as CK is not my main focus anymore. I'll try to have a look at it.14:33
huatszhangguoqing: I meant to have a decision on the integration :)14:33
huatsliujiong: sorry for the time to answer14:33
aolwaszhangguoqing: I'll check with you next week and see how I can help14:33
huatsI definitly try to go through all pending reviews in the next couple of days14:33
zhangguoqingaolwas: thank you very much. :)14:34
huatsI think we are done with the BP.14:34
huatsI think we can move to the open discussion. Does anyone has some topipc to adress ? need some help ?14:35
liujiongWill we have a regular meeting time in the future?14:35
sheeprinehuats: Got any news from openstack marketing about new logo and material ?14:35
sheeprineliujiong: Slot was removed because it was unused14:36
sheeprineIf we keep having regular ones I think we can reclaim one14:36
sheeprineBut having meeting in our channel is fine too, except we don't get meetbot14:36
huatsliujiong: I was about to propose a next meeting early january14:37
liujiongOK14:37
huatsa regular meeting is quite a lot of effort and we ware taking a very demanded time frame. So we agreed to release it14:37
liujiongunderstood14:37
huatsand to have our meetings here14:37
zhangguoqingWe need more code review, many of mine are not review. :)14:38
huatsI propose that we have out next meeting on january 5th at 13:00 UTC. How does it sound ?14:38
*** ivanromanko has joined #cloudkitty14:39
huatsor do you prefer earlier ?14:39
liujiongI'm fine with that.14:39
huats(regarding your question sheeprine, nothing since the draft proposal I have sent on the -dev mailing list)14:39
zhangguoqinghuats: agree with you january 5th at 13:00 UTC.14:39
gpocentekhuats: works for me14:40
huatsok good !14:40
huatsThank you everyone for attending it. I'll send a recap on the dev mainling list with a reminder of the next meeting14:41
aolwasok for me14:41
liujiongthank you, huats.14:41
zhangguoqinganother problem: We need more code review, many of mine are not review. :)14:42
huatszhangguoqing: I agree14:43
huatsbut we are working on that :)14:43
huatsI think that openning our work, doing regular IRC meetings like that are key to get more reviewers14:43
huatsso let's continue the good job !!!14:44
liujiongyes, we need more communication.14:44
* huats has never seen so many people connected on that channel !14:44
zhangguoqingI think, cores should review every three days at least.14:44
huatszhangguoqing: I think we might be able to have some strng rules like that once we have many contributors :(14:46
liujiongwe got a small group for now and focused.14:46
zhangguoqingGreat!14:46
liujiongI believe huats is also desired to do that.14:46
huatslof course14:47
zhangguoqinggood night.14:50
peschk_lgood night!14:51
liujionggood night, guoqing.14:51
*** zhangguoqing has quit IRC14:56
*** ivanromanko has quit IRC14:57
openstackgerritChristophe Sauthier proposed openstack/cloudkitty: Use keystone v3 instead of keystone v2 in cloudkitty's devstack plugin.  https://review.openstack.org/40981714:59
*** zhangguoqing has joined #cloudkitty15:01
*** zhangguoqing has quit IRC15:04
openstackgerritMerged openstack/cloudkitty: Pin kombu to < 4.0.0 to fix gate error  https://review.openstack.org/40905215:59
openstackgerritJeremy Liu proposed openstack/cloudkitty: Introduce hacking check to Cloudkitty  https://review.openstack.org/39693316:05
openstackgerritJeremy Liu proposed openstack/cloudkitty: Remove discover from test-requirements  https://review.openstack.org/39055316:07
openstackgerritChristophe Sauthier proposed openstack/cloudkitty: Use keystone v3 instead of keystone v2 in cloudkitty's devstack plugin  https://review.openstack.org/40981716:14
*** liujiong has quit IRC16:54
*** peschk_l has quit IRC17:05
*** sacharya has joined #cloudkitty17:06
*** openstack has joined #cloudkitty17:58
*** sacharya_ has joined #cloudkitty18:21
*** sacharya has quit IRC18:23
*** dinghh has quit IRC20:01
*** jwcroppe has quit IRC21:01
*** sacharya_ has quit IRC21:12
*** sacharya has joined #cloudkitty21:12
*** jwcroppe has joined #cloudkitty23:10

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!