Tuesday, 2015-06-02

« Monday, 2015-06-01 Index Wednesday, 2015-06-03 »
*** JAHoagie has quit IRC00:23
*** flwang1 has quit IRC00:37
*** flwang has joined #openstack-zaqar00:51
*** amitgandhinz has joined #openstack-zaqar00:53
*** cknight has joined #openstack-zaqar01:21
*** cknight has quit IRC01:21
*** cknight has joined #openstack-zaqar01:22
*** boris-42 has quit IRC01:35
*** Yanyanhu has joined #openstack-zaqar01:38
*** shakamunyi_ has joined #openstack-zaqar01:46
*** shakamunyi_ has quit IRC01:46
*** boris-42 has joined #openstack-zaqar01:55
*** cknight has quit IRC02:00
*** amitgandhinz has quit IRC02:18
*** amitgandhinz has joined #openstack-zaqar02:19
*** shakamunyi has joined #openstack-zaqar02:26
*** shakamunyi has quit IRC02:33
*** amitgandhinz has quit IRC03:19
*** asalkeld has joined #openstack-zaqar03:34
asalkeldhi, does zaqar have aggregation? so make a queue and please copy messages from these 3 other queues into it03:34
*** exploreshaifali has joined #openstack-zaqar04:26
*** ryansb has quit IRC04:41
*** JAHoagie has joined #openstack-zaqar04:48
*** ryansb has joined #openstack-zaqar04:49
*** ryansb has joined #openstack-zaqar04:49
*** exploreshaifali has quit IRC04:51
*** JAHoagie has quit IRC05:13
*** JAHoagie has joined #openstack-zaqar05:14
*** ryansb_ has joined #openstack-zaqar05:22
*** ryansb_ has joined #openstack-zaqar05:22
*** ryansb has quit IRC05:24
*** ryansb_ is now known as ryansb05:24
*** AAzza has quit IRC05:25
*** AAzza has joined #openstack-zaqar05:27
*** achanda has joined #openstack-zaqar05:30
*** reed has quit IRC05:33
*** AAzza has quit IRC05:34
*** AAzza has joined #openstack-zaqar05:36
*** achanda_ has joined #openstack-zaqar05:37
*** achanda has quit IRC05:38
*** reed has joined #openstack-zaqar05:47
openstackgerritOpenStack Proposal Bot proposed openstack/zaqar: Imported Translations from Transifex  https://review.openstack.org/18627106:04
*** AAzza has quit IRC06:05
*** AAzza has joined #openstack-zaqar06:05
*** achanda_ has quit IRC06:07
therveasalkeld, No07:43
*** dguerri`away is now known as dguerri07:48
*** dynarro has joined #openstack-zaqar07:48
*** vkmc has quit IRC08:45
*** vkmc has joined #openstack-zaqar08:45
*** dynarro has quit IRC08:45
*** achanda has joined #openstack-zaqar08:54
*** achanda_ has joined #openstack-zaqar08:56
*** achanda has quit IRC09:00
*** achanda_ has quit IRC09:21
asalkeldthanks therve , just checking09:34
vkmcasalkeld, hi! it should be impossible to implement those... considering how we store messages09:37
vkmcbut we should add some constrains to it probably09:38
asalkeldvkmc: ok, thanks09:38
vkmcsomething to think about in the future maybe09:38
asalkeldjust thinking about the cross project event stuff: https://review.openstack.org/#/c/185822/09:38
therveasalkeld, That last response was a bit confusing09:44
asalkelddoh09:44
therveAlso you didn't really answer Zane's main question AFAICT09:45
therveasalkeld, FWIW I believe webhooks are somewhat covered by ceilometer alarms09:46
asalkeldtherve: so i was basically liking the overall architecture of http://blog.caplin.com/2011/01/13/pubsubhubbub-vs-client-push-technologies-whats-the-difference/09:46
asalkeldseparate storage, hub, aggregator (at  least logically)09:47
asalkeldand was trying to argue down to what we still need09:47
asalkeldso the very last "take 3" is what i am suggesting (if we use zaqar)09:49
asalkeldI am a little worried about using features of zaqar that people are concerned about (listing, random access) unless it support a thing that is not a queue09:50
therveAgain, I'm more worried about clarifying what you actually want to achieve with that spec09:51
therveInstead of the technology09:51
flaper87It's not listing if you think of it as consuming09:51
flaper87the "listing" bit is a misconception introduced by previous reviews09:51
flaper87You pop messages out of the queue09:51
flaper87you don't need to list them09:51
asalkeldtherve: the problem statement in the spec is still valid afaik09:52
flaper87if you want to be protected from consumer's failures, then you'd probably claim them and get sets of messages09:52
flaper87sorry if I'm not participating much in the convo, it's a holiday here09:52
therveasalkeld, So applications or people?09:52
asalkeldtherve: we need to solve for both09:53
asalkeldbut happy to put the "people" to later, and solve the "application" first09:53
asalkeldflaper87: i like queues, just part of the features i need look and smell more like a feed/inbox09:54
asalkeldso don't want to try and force a circle into a sqare09:54
asalkeldsquare09:54
asalkeldtherve: zaqar already has webhook notifications09:57
therveIt does09:58
flaper87asalkeld: well, it kinda behaves more like a feed/inbox - queue's are more like containers for messages and not all the queue's guarantees exist in Zaqar - but just because it has extra semantics, it doesn't mean you need to use them. I'm talking about random access to messages.09:58
flaper87I need to read that review09:58
asalkeldflaper87: the initial post was a minimal solution, read the comments from zane and I09:59
flaper87asalkeld: kk, thanks09:59
asalkeldflaper87: happy lean on zaqar features, and let us know if any of those suggested features/ideas are crazy10:00
*** achanda has joined #openstack-zaqar10:00
*** Yanyanhu has quit IRC10:00
flaper87asalkeld: sure thing, I'll probably do it tomorrow, though10:00
flaper87but I'll definitely do it10:00
asalkeldthx10:00
asalkeldvkmc: nice -> https://github.com/openstack/zaqar/commit/adab329bf584e7fd5cc71b25e885907d50ea81bf10:10
asalkeldvkmc: any client side examples/docs on how to use that?10:11
flaper87asalkeld: it's not complete yet. dynarro will be working on that during Liberty10:11
flaper87there are still a few bits of the API missing10:12
flaper87unfortunately, Kilo was not enough to make that happen10:12
flaper87:(10:12
asalkeldok, but looks promising10:12
flaper87but vkmc is still working on the server side10:12
asalkeldi have also been experimenting with autobahn10:12
flaper87oh yeah, next step is to have a JS example running on the browser10:13
asalkeldlooks quite nice, but super annoying that every library uses a different mainloop10:13
vkmcasalkeld, :D10:13
vkmcI created a JS example some time ago... I should look for it10:14
flaper87asalkeld: right?10:14
flaper87I was sooooo annoyed about that10:14
flaper87I wish I had enough time to hack something that doesn't have its own mainloop10:14
asalkeldflaper87: yeah nuts, this whole mainloop stuff in python is very splintering10:15
asalkeldquite destructive i think10:15
asalkeldflaper87: a python example would be nice too10:16
*** dguerri is now known as dguerri`away10:37
therveRegarding loops, just use asyncio or something compatible with it10:54
*** achanda_ has joined #openstack-zaqar10:56
*** achanda has quit IRC10:56
*** achanda_ has quit IRC11:28
*** dynarro has joined #openstack-zaqar11:42
flaper87therve: autobah has support for asyncio, IIRC11:54
vkmcasyncio and twisted, yes11:55
*** achanda has joined #openstack-zaqar12:09
*** achanda has quit IRC12:20
*** achanda has joined #openstack-zaqar12:21
*** achanda has quit IRC12:25
*** dynarro has quit IRC12:53
*** dynarro has joined #openstack-zaqar13:06
*** sriram has joined #openstack-zaqar13:07
openstackgerritThomas Herve proposed openstack/zaqar: Add cleanups to several tests  https://review.openstack.org/18757313:17
*** dguerri`away is now known as dguerri13:31
*** amitgandhinz has joined #openstack-zaqar13:36
*** mpanetta has joined #openstack-zaqar13:37
*** malini has joined #openstack-zaqar13:45
*** cknight has joined #openstack-zaqar13:58
*** csoukup has joined #openstack-zaqar14:00
*** malini has quit IRC14:00
*** malini has joined #openstack-zaqar14:05
*** malini has quit IRC14:13
*** dynarro has joined #openstack-zaqar14:13
*** JAHoagie has quit IRC14:15
*** boris-42 has quit IRC14:18
*** malini has joined #openstack-zaqar14:20
openstackgerritThomas Herve proposed openstack/zaqar: Speed up message expiration tests  https://review.openstack.org/18760814:31
*** malini has quit IRC14:32
*** malini has joined #openstack-zaqar14:41
*** kgriffs|afk is now known as kgriffs14:50
*** exploreshaifali has joined #openstack-zaqar14:51
*** exploreshaifali has quit IRC14:52
*** kgriffs is now known as kgriffs|afk15:01
*** ametts has joined #openstack-zaqar15:19
*** barra204 has joined #openstack-zaqar15:35
*** shakamunyi has joined #openstack-zaqar15:35
*** boris-42 has joined #openstack-zaqar15:37
*** openstackgerrit has quit IRC15:42
*** openstackgerrit has joined #openstack-zaqar15:43
*** cknight has quit IRC15:57
*** amitgandhinz has quit IRC16:11
*** dynarro has quit IRC16:16
*** mpanetta has quit IRC16:28
*** mpanetta has joined #openstack-zaqar16:52
*** cknight has joined #openstack-zaqar16:53
*** dynarro has joined #openstack-zaqar17:00
*** malini has quit IRC17:02
*** dguerri is now known as dguerri`away17:04
*** achanda has joined #openstack-zaqar17:08
*** malini has joined #openstack-zaqar17:09
*** malini has quit IRC17:13
*** malini has joined #openstack-zaqar17:13
*** malini has quit IRC17:25
*** amitgandhinz has joined #openstack-zaqar17:38
*** dynarro has quit IRC17:53
*** dynarro has joined #openstack-zaqar17:54
*** dynarro has quit IRC17:58
*** malini has joined #openstack-zaqar17:59
*** ametts has quit IRC17:59
*** openstackgerrit has quit IRC18:09
*** openstackgerrit has joined #openstack-zaqar18:10
*** dguerri`away is now known as dguerri18:11
*** achanda has quit IRC18:45
*** openstackgerrit has quit IRC18:56
*** openstackgerrit has joined #openstack-zaqar18:56
*** achanda has joined #openstack-zaqar19:03
*** achanda has quit IRC19:26
*** mpanetta has quit IRC19:51
*** malini has quit IRC20:07
*** openstackgerrit has quit IRC20:10
*** openstackgerrit has joined #openstack-zaqar20:10
openstackgerritRyan Brown proposed openstack/zaqar: s/Marconi/Zaqar/ in method docstring  https://review.openstack.org/18775520:20
openstackgerritRyan Brown proposed openstack/zaqar: Clarify code comment in Redis backend  https://review.openstack.org/18775620:20
*** kgriffs|afk is now known as kgriffs20:26
*** kgriffs is now known as kgriffs|afk20:29
flwangvkmc: ping21:02
*** shakamunyi_ has joined #openstack-zaqar21:03
*** shakamunyi__ has joined #openstack-zaqar21:04
*** shakamunyi__ is now known as tylerdurden21:05
*** tylerdurden has quit IRC21:08
*** shakamunyi_ has quit IRC21:08
*** sriram has quit IRC21:09
flaper87flwang: hey there21:12
flaper87flwang: glad to see you around21:12
flaper87did you get my messages?21:12
flwangflaper87: which one? i think i missed21:12
flaper87flwang: np, we discussed specs last night21:12
flaper87(during the meeting YOU MISSED ;)21:12
flwangoh21:12
flwangyou mean the meeting is last night?21:12
flaper87I think we all love your spec but I nitpicked on some things21:13
flwangwhat's your current time?21:13
flaper87well, it was 9am Tuesday for you21:13
flwangah, shit21:13
flaper87flwang: right21:13
flaper87so, we'll have it at the same time next week21:13
flaper87that is: Monday 21 UTC21:13
flwangi was taking a sick leave yesterday since my son and I got bad caughing21:13
flaper87be there21:13
flwangsorry about that21:13
flaper87flwang: no worries, I figured something serious was going on21:14
flwangreally sorry21:14
flaper87that said, we've got some good feedback21:14
flaper87I'd love your input on the specs21:14
flaper87and I'd appreciate if you could dedicate some time to reviews21:14
flwangsure, I will21:15
flwangflaper87: so what's the focus of us in L?21:15
flaper87flwang: integration with other projects. Either contributing to them or implementing things that make sense to have in Zaqar and would make their lives simpler21:16
flaper87That'd be:21:16
flaper871) pre-signed urls21:16
flaper872) policies21:16
flaper873) websockets21:16
flaper874) client library up-to-speed21:16
flaper87Heat folks are doing work already on their side21:17
flaper87something I don't want us to foresee is deployment21:17
flaper87I'd like to take Ryan's (the puppet guy) word and encourage him to contribute to puppet-zaqar21:17
flwangflaper87: cool, questions are coming21:17
* flaper87 listens carefully21:17
flwang1. as for the user event, is it depending on notifications or the basic 'queue' service?21:17
flaper87both ?21:18
flaper87Most of it is the messaging API21:19
flaper87but the notification API would be very useful21:19
flwangI know we difinitely depend on the basic queue service, my point is21:19
flaper87It'd be great to have some folks contribute to other pushing mechanisms21:19
openstackgerritMerged openstack/zaqar: s/Marconi/Zaqar/ in method docstring  https://review.openstack.org/18775521:19
flwangif the user event depends on the new notification service, we may need put some effort on that, since it's really young21:19
flwangflaper87: yep, I would like to work out a email driver and sms driver in L21:20
flaper87flwang: oh yeah21:20
flaper87flwang: how long do you think it'll take you to do policy?21:20
flwangbut if there are any other folk can give us a hand21:20
flaper87I think L-1 is a good milestone21:20
flwangI would like to focus on the framework21:20
vkmcflaper87, flwang o/21:20
flwangto make it more table21:20
flwangstable21:20
flaper87flwang: do you want me to take policy so you can focus on notifications ?21:21
flwangflaper87: i don't mind21:21
flwangyou are the boss :)21:21
flaper87I'm fine as long as you keep patches coming :)21:21
flaper87you might need to write a couple of new specs though21:22
flwangflaper87: one thing i'm curious is21:22
flaper87with the ideas/improvements you have in mind21:22
flwangwhy kgriffs|afk didn't mention the policy thing given Rackspace is using it21:22
flwangi mean if there are strong requirement from customer about that, though I think it's very important21:22
flaper87We've talked about policy21:23
flwangok21:23
flaper87we just haven't had the time to get back to it21:23
flaper87I think it's useful and I've also some ideas that'll require it21:23
flaper87like per-queue policies21:23
flwangflaper87: i like per queue polices, it can be implemented by leveraging the queue metadata i think21:24
flwangit would be interesting21:24
flaper87flwang: right but to do that we need private fields in the metadata21:24
flaper87which is something I'm writing down21:24
flaper87but to have private fields, we need policies21:25
flaper87:P21:25
flaper87or not21:25
flaper87but anyway, something I've been brewing in my head21:26
flaper87I'll hopefully have a spec tomorrow21:26
flwangah, so your per-queue policy will depend on the basic policy, right?21:27
flaper87exactly21:28
flwanggot it21:28
flaper87but you'd store the per-queue policy in a private field inside the queue metadata21:28
flaper87and we don't have private fields21:28
flwangyes21:28
flaper87we need policies to protect those fields21:28
flwangjust like the protected properties in Glance21:29
flaper87anyway, I'll write it down21:29
flaper87flwang: right21:29
flaper87flwang: btw, can you update the policy spec so I can approve it?21:29
flwangyep, I will update it21:29
flwangbtw, what did you mean missing the request context?21:29
flaper87flwang: do we need this? https://github.com/openstack/glance/blob/master/glance/api/middleware/context.py21:31
flaper87and oslo.context21:31
flwangdid you see this? https://review.openstack.org/#/c/182156/21:31
flaper87actually, I've no idea if we have that already21:31
flaper87flwang: <321:31
flaper87nothing else to add21:31
flwangI think i'm on the way :)21:31
flaper87good, good, gooooooooooooooooooooooood21:32
flwangflaper87: so we're going to land it in L-1, right?21:32
* flaper87 is super excited21:32
flaper87flwang: policies? yes21:32
flaper87flwang: I'd be super cool to get the specs for your plans on notifications up for review this week21:33
flwangok, I'm updating the spec and will throw a patch set in 1 min21:33
flaper87awesome, thank you!21:33
flwangflaper87: oh man, you're always pushy like a crazy boss21:33
flwangbut i like it21:33
flaper87flwang: hahahaha, that's a first but I'll take it :D21:34
flaper87WE NEED TO CATCH UP!!!!!!!!!!121:34
*** malini has joined #openstack-zaqar21:35
flwangvkmc: please join the chat :)21:38
flaper87flwang: vkmc is living what we Europeans live!21:38
flaper87let her suffer in silence21:38
flaper87:P21:38
flwanglol21:39
flwangflaper87: btw, as for the notifications driver, should I create 2 specs for email and sms driver?21:39
flaper87flwang: yeah, I'd say so21:40
flaper87sorry :(21:40
flaper87thing is they are quite different21:40
flwangno worries, will do that21:40
flwangas for the framework improvement, I just have some pieces of ideas, I need to aggregate them21:41
openstackgerritFei Long Wang proposed openstack/zaqar-specs: Policy support  https://review.openstack.org/17967321:42
*** dguerri is now known as dguerri`away21:51
flwangflaper87: vkmc: https://review.openstack.org/#/c/179673/ and https://review.openstack.org/#/c/182156/ thanks :)21:59
*** amitgandhinz has quit IRC22:00
kragnizflaper87: yo!22:00
kragnizflaper87: reviews pls https://review.openstack.org/#/c/186099/22:00
*** cknight has quit IRC22:03
flaper87kragniz: <322:04
kragnizflaper87: <322:05
flaper87kragniz: done22:05
kragnizty!22:05
*** malini has quit IRC22:06
kragnizflaper87: are those sort of reviews useful?22:08
flaper87kragniz: you mean the patch you proposed ?22:08
flaper87absolutely22:08
kragnizcool22:08
flaper87however, if you'd like to do something more interesting let me know22:08
flaper87I'm happy to help you find something that won't make Stuart kill you22:09
flaper87or me22:09
kragnizI sometimes wonder about trivial things like that22:09
kragnizflaper87: lol22:09
flwangkragniz: i think a notifications driver is the one :)22:13
kragnizflwang: eek that sounds scary! :)22:18
openstackgerritMerged openstack/zaqar: Port to oslo_log  https://review.openstack.org/18609922:18
flaper87kragniz: ^22:21
kragnizflaper87: <322:22
vkmchey hey hey22:24
vkmcI was having dinner22:24
vkmc:D22:24
* vkmc reads backlog 22:24
flaper87vkmc: dafuq ?22:25
flaper87I hope you're referring to the *SECOND*22:25
flaper87dinner22:26
vkmcof course22:26
vkmcthere are two dinners22:26
vkmcalways22:26
flaper87ok ok22:26
flwangflaper87: so do we have assignees for all of those priorities in L?22:29
flaper87yup, we do22:29
flaper87flwang: that'd be you22:29
flaper87:D22:29
flaper87kragniz: want to add oslo.policy support to Zaqar ?22:29
flaper87:)22:29
flaper87That should be pretty straightforward22:29
flaper87kragniz: and flwang wrote the spec for you22:30
flaper87:P22:30
flaper87I can provide candies22:30
flaper87and jaggerbombs22:30
kragnizflaper87: sure thing22:30
flaper87kragniz: awesome!22:30
flaper87flwang: assign kragniz to the policy spec22:30
flaper87(before he changes his mind)22:30
flwangkragniz: are you happy to take it? :)22:31
kragnizsure!22:31
flwangawesome22:31
kragnizI'll need to find a bit of time for it22:31
flwangcome on22:31
kragnizbut that's what the time you could be sleeping is for22:32
flwangwe just wrote down we need to complete it in L-122:32
flaper87kragniz: it won't take you much, really.22:32
kragnizcool22:32
kragnizI'm up for it22:32
* flaper87 ducks22:32
flaper87ok, I'm off22:33
* kragniz gooses22:33
flwangflaper87: pls bless it https://review.openstack.org/#/c/179673/22:33
kragnizflwang: nn22:33
flaper87today was supposed to be a holiday22:33
flwangkragniz: ^22:33
flaper87flwang: ok22:33
flaper87flwang: you need to change the assignee22:34
flwangshit :)22:34
flaper87I'll give it another read tomorrow and approve it :)22:34
flaper87off to read some papers22:34
flwangflaper87: you win22:34
flaper87flwang: you need to learn something about me: I either win or tie ;)22:34
* flaper87 stops throwing bullshit in this amazing channel and goes for the night22:35
flwanghave a good night22:35
openstackgerritLouis Taylor proposed openstack/zaqar-specs: Policy support  https://review.openstack.org/17967322:44
kragnizflwang: I updated the spec ^22:44
flwangkragniz: ok, cool22:45
*** amitgandhinz has joined #openstack-zaqar23:11
*** amitgandhinz has quit IRC23:15
*** csoukup has quit IRC23:18
*** bradjones has quit IRC23:43
*** bradjones has joined #openstack-zaqar23:45
*** diga_ has joined #openstack-zaqar23:46
*** yfujioka has joined #openstack-zaqar23:55
*** barra204 has quit IRC23:57
*** shakamunyi has quit IRC23:57
*** diga_ has quit IRC23:59
« Monday, 2015-06-01 Index Wednesday, 2015-06-03 »

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