Thursday, 2016-01-07

« Wednesday, 2016-01-06 Index Friday, 2016-01-08 »
*** kgriffs is now known as kgriffs|afk00:00
*** kgriffs|afk is now known as kgriffs00:01
*** kgriffs is now known as kgriffs|afk00:11
Eva-iflwang: can you please confirm this nasty bug: https://bugs.launchpad.net/zaqar/+bug/1531671 ?00:41
openstackLaunchpad bug 1531671 in zaqar "Websocket html client example can't process notifications" [Undecided,New] - Assigned to Eva Balycheva (ubershy)00:41
*** boris-42 has quit IRC01:13
flwangEva-i: nice catch01:14
Eva-iflwang: thanks =)01:18
flwangEva-i: i'm really happy to have you in the team, i couldn't help tell you the secret01:27
Eva-iflwang: oh01:40
Eva-iflwang: I feel the same about you. I think you are wise and kind to me. I never had a team leader before, but I think other team leaders don't use virtual hugs to motivate team members, so I'm lucky01:50
Eva-i=)01:50
flwang:D01:51
*** flwang1 has quit IRC01:58
*** flwang1 has joined #openstack-zaqar01:58
openstackgerritEva Balycheva proposed openstack/zaqar: Make websocket.html process notifications  https://review.openstack.org/26451302:01
*** davideagnello has quit IRC02:07
*** agileclipse has joined #openstack-zaqar02:54
*** boris-42 has joined #openstack-zaqar03:02
*** agileclipse has quit IRC03:09
*** mdnadeem has joined #openstack-zaqar03:16
*** mdnadeem has quit IRC03:20
*** mdnadeem has joined #openstack-zaqar03:21
*** agileclipse has joined #openstack-zaqar03:21
openstackgerritLiuNanke proposed openstack/zaqar: Make enforce_type=True in CONF.set_override  https://review.openstack.org/26335903:59
*** agileclipse has quit IRC04:05
*** flwang1 has quit IRC04:06
*** flwang1 has joined #openstack-zaqar04:08
*** flwang1 has quit IRC04:16
*** flwang1 has joined #openstack-zaqar04:30
*** mdnadeem has quit IRC05:00
*** stanchan has joined #openstack-zaqar05:07
*** GB21 has joined #openstack-zaqar05:38
*** mdnadeem_ has joined #openstack-zaqar06:12
openstackgerritOpenStack Proposal Bot proposed openstack/zaqar: Updated from global requirements  https://review.openstack.org/26447206:58
*** GB21 has quit IRC08:01
*** dynarro has joined #openstack-zaqar08:41
*** GB21 has joined #openstack-zaqar08:59
*** GB21 has quit IRC09:30
*** GB21 has joined #openstack-zaqar09:31
*** stanchan has quit IRC09:45
openstackgerritMD NADEEM proposed openstack/python-zaqarclient: Flavor create CLI support for v2  https://review.openstack.org/25024809:49
*** GB21 has quit IRC09:54
*** GB21 has joined #openstack-zaqar09:57
openstackgerritMD NADEEM proposed openstack/python-zaqarclient: Query claim CLI support for v1  https://review.openstack.org/25880109:58
openstackgerritMD NADEEM proposed openstack/python-zaqarclient: Claim renew CLI support for v1  https://review.openstack.org/26092710:01
openstackgerritMD NADEEM proposed openstack/python-zaqarclient: Query claim CLI support for v1  https://review.openstack.org/25880110:05
*** mdnadeem_ has quit IRC10:09
*** mdnadeem has joined #openstack-zaqar10:19
*** GB21 has quit IRC10:50
*** GB21 has joined #openstack-zaqar10:53
*** dynarro has quit IRC10:53
*** GB21 has quit IRC11:00
*** GB21 has joined #openstack-zaqar11:02
*** GB21 has quit IRC11:10
*** GB21 has joined #openstack-zaqar11:12
*** GB21 has quit IRC11:16
mdnadeemflwang, flwang1: Need your help to finalyzed it >> https://review.openstack.org/#/c/246147/611:31
mdnadeemPlease have a look ^^11:31
*** GB21 has joined #openstack-zaqar11:35
*** GB21 has quit IRC12:15
*** godlike has quit IRC13:19
*** godlike has joined #openstack-zaqar13:21
*** godlike has quit IRC13:21
*** godlike has joined #openstack-zaqar13:21
thervemdnadeem, ryansb: Around? I have a question about https://review.openstack.org/#/c/245526/13:48
therveIt seems that it breaks compatibility (os_auth_token changed to auth_token), am I missing something?13:48
ryansbtherve: looking14:01
*** njohnsto_ has joined #openstack-zaqar14:03
*** GB21 has joined #openstack-zaqar14:04
openstackgerritEva Balycheva proposed openstack/zaqar: Refactoring of docs during Mitaka cycle  https://review.openstack.org/25058214:05
*** njohnsto_ has quit IRC14:13
*** njohnsto_ has joined #openstack-zaqar14:14
*** njohnsto_ has quit IRC14:16
*** njohnsto_ has joined #openstack-zaqar14:16
*** amitgandhinz has joined #openstack-zaqar14:17
*** sriram has joined #openstack-zaqar14:27
*** amitgandhinz has quit IRC14:29
*** amitgandhinz has joined #openstack-zaqar14:29
openstackgerritEva Balycheva proposed openstack/zaqar: Make websocket.html process notifications  https://review.openstack.org/26451314:41
therveryansb, So wdyt?14:52
*** GB21 has quit IRC15:01
*** malini has joined #openstack-zaqar15:01
*** pradk has joined #openstack-zaqar15:01
*** GB21 has joined #openstack-zaqar15:01
ryansbhm, you're right that is likely to be a compatibility problem15:03
ryansb(assuming your refer to openstack/python-zaqarclient/zaqarclient/auth/keystone.py L65)15:03
ryansbs/your/you/15:03
*** ametts has joined #openstack-zaqar15:06
therveryansb, Yeah I do15:07
therveThat's okay, I'll submit a little patch15:07
therveIt should be in before the next release15:07
ryansbyeah, we're not even in M2 until next Monday I think15:08
ryansberr, I mean we don't finish M215:08
openstackgerritThomas Herve proposed openstack/python-zaqarclient: Handle os_auth_token option in keystone config  https://review.openstack.org/26482715:13
therveclient release are somewhat independent anyway15:13
Eva-iryansb: therve: hello15:17
therveEva-i, Hi15:17
ryansbmorning folks15:18
ryansbtherve: how would you feel about: return self.conf.get(k, self.conf.get("os_%s" % k))15:20
Eva-iryansb: therve: how do you think, is it okay that subscription notifications are so concise, i.e. are lacking some important info (IMO), for example. name of the queue or subscription id/message id. Now notifications look kind of like this: {"body": "some message", "ttl": 300}15:20
ryansbinstead of the oneline if15:20
therveryansb, Shrugs, I just modified the existing code15:21
therveEva-i, It's okay that they are concise. It doesn't mean they could contain more info15:23
therve*couldn't*15:24
ryansbIMO a message ID would be a really good idea to include15:25
ryansbthe queue name less so, since consumers know what they're suscribed to15:26
Eva-itherve: imagine a use case, when the user want to subscribe to few queues. Zaqar server will send notifications and it would be impossible to figure out from which queue the message came.15:26
Eva-itherve: of course the user can subscribe to different listening ports, but it's not very convenient15:27
therveEva-i, Feel free to improve them15:28
therveMy point is that it's a missing feature, not a bug15:28
Eva-ioh, my speech is not very good now, but I hope you understood what I said15:29
Eva-itherve: okay15:29
ryansbcompare to http://docs.aws.amazon.com/sns/latest/dg/json-formats.html#http-notification-json which includes both topic and message ID15:29
ryansbEva-i: IMO it would be good to add them15:29
ryansbfortunately, that's not a breaking change since it would just be adding JSON keys15:30
therveYeah it's easy to do it in a compatible way15:30
Eva-iryansb: it's not a breaking change also according to the fact that api v2 is not stable, we can do many things we want15:30
Eva-iShould I propose a spec?15:31
*** dynarro has joined #openstack-zaqar15:31
openstackgerritThomas Herve proposed openstack/python-zaqarclient: Handle os_auth_token option in keystone config  https://review.openstack.org/26482715:32
*** kgriffs|afk is now known as kgriffs15:33
ryansbEva-i: well that's true too. You can propose a spec. IMO you don't have to make a huge one since the spec will basically just need to say15:33
ryansb1) SNS includes ID and topic in notification15:34
ryansb2) subscribers are likely to want queue information if they subscribe to multiple queues15:34
ryansb3) subscribers that want to ack/delete/claim messages they are notified of will need the message ID15:34
ryansbplease add me as a reviewer whenever you propose it, or I can write it up if you're short on time15:35
Eva-iryansb: oh, Ryan, you wrapped my thoughts into nice words, thank you. =)15:37
Eva-iryansb: okay, I'll add you as a reviewer when I propose a spec/patch. But before I think I should ask also flwang. He's not very enthusiastic on adding many features during this cycle.15:38
Eva-itherve: thank you too for your opinion15:39
therveSure15:40
ryansbkk - we'll see how it works out15:40
Eva-iI have a newbie problem. This bug is assigned to me https://bugs.launchpad.net/python-zaqarclient/+bug/1529846, I proposed a change and it was merged. But I forgot to add "Closes-Bug:" tag to the commit message. Should I manually close this bug?15:47
openstackLaunchpad bug 1529846 in Python client library for Zaqar "CLI fails to parse v2 pool list entry point" [High,Confirmed] - Assigned to Eva Balycheva (ubershy)15:47
ryansbEva-i: yep15:48
Eva-iryansb: oki15:52
*** csoukup has joined #openstack-zaqar16:18
*** Eva-i has quit IRC16:32
*** Eva-i has joined #openstack-zaqar16:33
*** mdnadeem has quit IRC17:02
*** GB21 has quit IRC17:11
*** davideagnello has joined #openstack-zaqar17:15
openstackgerritOpenStack Proposal Bot proposed openstack/zaqar: Updated from global requirements  https://review.openstack.org/26447217:19
*** kgriffs is now known as kgriffs|afk17:59
*** dynarro has quit IRC18:16
*** malini has quit IRC18:40
*** pradk has quit IRC18:47
*** malini has joined #openstack-zaqar19:09
*** pradk has joined #openstack-zaqar19:14
*** kgriffs|afk is now known as kgriffs19:16
*** c_soukup has joined #openstack-zaqar19:29
*** ameade_ has joined #openstack-zaqar19:35
openstackgerritMerged openstack/python-zaqarclient: Claim create CLI support for v1  https://review.openstack.org/25838119:36
*** cpallare- has joined #openstack-zaqar19:37
*** bradjones_ has joined #openstack-zaqar19:37
*** bradjones_ has joined #openstack-zaqar19:37
*** pradk has quit IRC19:38
*** csoukup has quit IRC19:38
*** bradjones has quit IRC19:38
*** ameade has quit IRC19:38
*** cpallares has quit IRC19:38
*** cpallare- is now known as cpallares19:38
*** bradjones_ is now known as bradjones19:38
*** pradk has joined #openstack-zaqar19:40
*** ameade_ is now known as ameade19:40
*** c_soukup has quit IRC20:12
*** openstackgerrit has quit IRC20:23
*** openstackgerrit has joined #openstack-zaqar20:24
openstackgerritMerged openstack/zaqar: Move subscription test from v1_1 to v2  https://review.openstack.org/26392020:33
*** malini has left #openstack-zaqar20:41
*** c_soukup has joined #openstack-zaqar20:50
Eva-iAm only I facing a problem when most of the zaqar client functional tests are failing?20:51
Eva-iflwang: flwang1: ping20:55
vkmcEva-i, hey there!20:56
vkmcthe functional tests in the gate are working AFAIK20:56
Eva-ivkmc: Hello, Vic20:56
vkmcEva-i, except flwang latest patches20:57
Eva-iI try to run functional tests on master branch, but there is failing20:57
*** kgriffs is now known as kgriffs|afk21:00
*** kgriffs|afk is now known as kgriffs21:02
*** kgriffs is now known as kgriffs|afk21:02
Eva-ivkmc: thanks for reviews =)21:04
vkmcEva-i, np :) sorry those were not done earlier21:05
vkmcEva-i, btw, I also had to mention that the spec for the websocket binary support looked awesome21:06
vkmcs/looked/looks21:06
Eva-ivkmc: hehe. I think now this spec is too big, especially "alternatives" section. =) Also ryansb asked me to make my next upcoming Spec smaller, I think it's not a coincidence. =)21:10
Eva-ivkmc: thanks21:10
*** kgriffs|afk is now known as kgriffs21:11
ryansbEva-i: it actually was - when we were talking I meant "the subscription format change is small, and we've discussed it pretty thoroughly, so no need to spend loads of time on a spec"21:11
ryansbthe binary spec is great :)21:12
vkmcEva-i, haha I think it's good to have all the information you need... more if you thought about that and you feel it is worth to mention it21:12
ryansbvkmc: have a stroupwaffle (#) you've merged a ton of changes today :)21:12
Eva-iryansb: alright21:13
* vkmc gets the stroopwafel (#) :D 21:15
vkmcthanks ryansb21:15
*** njohnsto_ has quit IRC21:29
*** mdnadeem has joined #openstack-zaqar21:30
*** njohnsto_ has joined #openstack-zaqar21:30
openstackgerritMerged openstack/zaqar: Make websocket.html process notifications  https://review.openstack.org/26451321:31
openstackgerritMerged openstack/zaqar: Refactoring of docs during Mitaka cycle  https://review.openstack.org/25058221:32
*** JAHoagie has joined #openstack-zaqar21:32
openstackgerritMerged openstack/zaqar: Updated from global requirements  https://review.openstack.org/26447221:32
openstackgerritMerged openstack/zaqar: delete python bytecode before every test run  https://review.openstack.org/25218521:32
ryansbmerge festival!21:36
*** njohnsto_ has quit IRC21:37
*** njohnsto_ has joined #openstack-zaqar21:38
*** njohnsto_ has quit IRC21:40
*** njohnsto_ has joined #openstack-zaqar21:40
*** mdnadeem has quit IRC21:43
*** godlike has quit IRC21:50
*** boris-42 has quit IRC21:50
*** godlike has joined #openstack-zaqar21:51
*** boris-42 has joined #openstack-zaqar21:51
*** amitgandhinz has quit IRC21:53
*** pradk has quit IRC22:24
flwangryansb: vkmc: Eva-i: ping22:33
vkmcflwang, pong22:33
ryansbwhassup?22:33
flwangi'm thinking the comment at https://review.openstack.org/#/c/249395/19/zaqarclient/queues/v2/subscription.py about the TTL value of subscription22:34
flwangnow the min TTL value for msg and claim is 60 secs22:34
flwangand I would like to use it for subscription as well. but that needs a change for zaqar server side22:34
flwangso i would like to know if you guys are happy if i just give the default value as 60 sec22:35
flwangfor the client patch and work out another patch for the server side22:35
flwangi'm keen to get the subscription patch in since it's blocking the ceilometer notifier integration22:36
*** njohnsto_ has quit IRC22:36
flwangand the ceilometer patch is on the way actually, see https://review.openstack.org/#/c/26430922:37
*** njohnsto_ has joined #openstack-zaqar22:37
flwangany comments?22:39
*** sriram has quit IRC22:40
vkmc++22:40
vkmcso... as the min ttl for claims and messages22:40
vkmcwe should set a configuration22:40
vkmcand a default value22:41
vkmcand use that value22:41
flwangthe TTL default value will be added with the fix of https://bugs.launchpad.net/zaqar/+bug/152916822:41
openstackLaunchpad bug 1529168 in zaqar "Zaqar should ensure deletion of expired subscriptions" [High,Confirmed] - Assigned to Fei Long Wang (flwang)22:41
vkmcin here I mean https://github.com/openstack/zaqar/blob/master/zaqar/transport/base.py#L30-L3722:41
flwanghttps://github.com/openstack/zaqar/blob/master/zaqar/transport/validation.py#L2422:42
flwangcurrently, it's hard code :(22:42
vkmcok, we could add it in a follow up22:42
vkmcand yeah, and I'll make them live longer than the messages and queues22:42
vkmcsince the subscriptions are expected to be used by several clients22:42
vkmcwhat do you guys thinks about that?22:43
flwangvkmc: i have figured out a way to get a permanent subscription22:43
flwangbut i haven't try it22:43
vkmcflwang, which way?22:43
flwangset the expires with a non-date format22:44
flwangsee https://github.com/openstack/zaqar/blob/master/zaqar/storage/mongodb/subscriptions.py#L9622:44
flwangby default, the 'expires' is a date format + TTL22:45
flwangi think we can do something like, 1. check if the TTL is -1 (if we use -1 as infinite),  2. if it's -1, then set the expires as None or anything else, but not a date format22:46
vkmcmakes sense22:46
vkmcnow + ttl expires22:46
vkmc-1 not expires22:46
flwang3. if TTL is int and > the min value(which has been verified by validation), then do it normally22:46
flwangbut as for redis, i haven't investigate how to achieve that22:47
vkmcwhy is it different for redis?22:47
vkmcoh, wait a minute22:47
vkmclet me check the code first22:47
vkmc:)22:47
vkmck, not easily visible22:48
vkmcnow yes22:48
vkmcwhy is it different?22:48
flwangwe're using a lua script to collect the expired ones, IIRC22:49
flwangi haven't reviewed the code, so i'm not sure if we can get a permanent subscription22:49
flwangfor redis22:50
flwangso are you guys ok for the client patch if we just use 60 as a default value and fix the TTL in a separate patch on the server side?22:53
vkmcI'm ok with that :)22:53
ryansb+122:53
flwangvkmc: is there any other concern for the subscription create patch?22:54
flwangi know the functional test failed, but it works on my local and Eva-i confirmed that22:54
vkmcflwang, I saw that the functional tests were failing22:54
vkmcnot sure if you check that out22:54
vkmcoh22:54
vkmcgood22:54
*** ametts has quit IRC22:55
flwangvkmc: yep, it's a little bit weird, i will investigate it later22:56
flwangand btw, if you guys are interested in, pls help review the aodh patch22:56
openstackgerritFei Long Wang proposed openstack/python-zaqarclient: Add support for subscription update v2  https://review.openstack.org/26130322:57
vkmcaodh?22:57
flwangit's the alarm project splitted from ceilometer :)22:58
flwangbig tent :D22:58
*** c_soukup has quit IRC22:58
openstackgerritFei Long Wang proposed openstack/python-zaqarclient: Add support for subscription create v2  https://review.openstack.org/24939522:58
flwangvkmc: ryansb: pls feel free bless it https://review.openstack.org/#/c/249395/ :)23:00
vkmcnice!23:00
vkmcyes yes23:00
vkmclet's wait for the gate23:00
vkmcand then move it fwd23:00
flwangvkmc: coooooooool23:00
flwangvkmc: ryansb: and there is another i would like to get your comments23:01
flwangas for the zaqar UI, i was trying to use anjular JS,  you know23:01
flwangbut i just found it's hard to use atm23:01
flwangso i'm wondering if we can use django just like most of the current panels didi23:02
flwangand migrate to anjular js later when most of the others panels migrated23:02
flwangdoes that make sense :)23:02
« Wednesday, 2016-01-06 Index Friday, 2016-01-08 »

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