| *** cpallares has quit IRC | 00:02 | |
| *** cpallares has joined #openstack-zaqar | 00:02 | |
| cpallares | boooooooooooooooooooooo | 00:43 |
|---|---|---|
| flwang | cpallares: hey, how are you | 00:44 |
| cpallares | flwang: hey :) I'm doing good and you? | 00:48 |
| cpallares | flwang: How's your little hacker? | 00:48 |
| flwang | i'm really good, it's the first time I didn't touch openstack more than 10 days | 00:49 |
| cpallares | flwang: oh wow that's a long time, which reminds me happy holidays and happy new years :) | 00:49 |
| flwang | my little hacker is good as well, still trying to interrupt me coding sometimes :) | 00:50 |
| flwang | cpallares: yep | 00:50 |
| cpallares | haha | 00:50 |
| cpallares | aww | 00:50 |
| cpallares | flwang: hey you're in New Zealand, right? Are you attending LinuxConfAu? | 00:51 |
| flwang | cpallares: yep, i'm in NZ, but i'm not sure if i will be LCA | 00:51 |
| cpallares | flwang: Aw you should go! I got a scholarship to attend from the python foundation. I want to attend the OpenStack ATC meetup there! There's going to be an openstack mini-conference. | 00:53 |
| flwang | cpallares: maybe, I need to talk with my manager to get some funding :D | 00:54 |
| cpallares | flwang: Do iiiiit! | 00:54 |
| cpallares | flwang: Also do it fast cause it's next week :P | 00:55 |
| flwang | cpallares: yep, i know | 00:55 |
| flwang | cpallares: the ticket is a little bit expensive | 00:57 |
| cpallares | flwang: Yeah, especially the professional one. | 00:58 |
| flwang | the ATC meetup is on 11th Jan, is it? | 00:59 |
| cpallares | flwang: Yes | 00:59 |
| cpallares | https://www.eventbrite.com.au/e/linuxconfau-2015-openstack-atc-meetup-tickets-14765282357 | 01:00 |
| cpallares | flwang: ^^ | 01:00 |
| cpallares | You don't need to be registered to attend. | 01:00 |
| flwang | cool, so if i just want to join the meetup, that may mean i only need an air ticket :) | 01:01 |
| cpallares | flwang: Yes :) | 01:03 |
| *** nakul_cpani has joined #openstack-zaqar | 01:03 | |
| cpallares | flwang: http://linux.conf.au/register/prices#registration-assistance | 01:05 |
| *** nakul_cpani has left #openstack-zaqar | 01:06 | |
| cpallares | If you send them a nice email, they might give you a nice discount on the registration. | 01:07 |
| flwang | cpallares: I'm not sure if i have much time to join the 5days conf | 01:08 |
| flwang | but i'm interested in the meetup on sunday | 01:08 |
| *** cpallares has quit IRC | 01:13 | |
| *** bradjones has quit IRC | 01:17 | |
| *** bradjones has joined #openstack-zaqar | 01:18 | |
| *** cpallares has joined #openstack-zaqar | 01:27 | |
| kragniz | cpallares: that python scholarship sounds great! | 01:45 |
| cpallares | hey kragniz :) | 01:59 |
| cpallares | kragniz: yeah, the python foundation gives awesome funding | 02:00 |
| cpallares | kragniz: I'm excited. I've heard great things about linuxconfau. It's the biggest linux grassroots conference. | 02:03 |
| kragniz | cpallares: cool! | 02:14 |
| *** echevemaster has quit IRC | 03:15 | |
| *** cpallares has quit IRC | 03:42 | |
| *** achanda has joined #openstack-zaqar | 04:46 | |
| *** nakul_cpani has joined #openstack-zaqar | 04:46 | |
| *** nakul_cpani has left #openstack-zaqar | 04:46 | |
| *** achanda has quit IRC | 05:20 | |
| *** achanda has joined #openstack-zaqar | 06:56 | |
| *** achanda has quit IRC | 07:44 | |
| *** achanda has joined #openstack-zaqar | 07:45 | |
| *** achanda_ has joined #openstack-zaqar | 07:47 | |
| *** achanda has quit IRC | 07:49 | |
| *** achanda_ has quit IRC | 07:52 | |
| *** achanda has joined #openstack-zaqar | 08:17 | |
| *** achanda has quit IRC | 08:59 | |
| *** achanda has joined #openstack-zaqar | 09:03 | |
| *** exploreshaifali has joined #openstack-zaqar | 09:09 | |
| *** achanda has quit IRC | 09:28 | |
| *** flwang1 has joined #openstack-zaqar | 09:52 | |
| flwang1 | flaper87: ping | 09:52 |
| *** shaifali_ has joined #openstack-zaqar | 10:16 | |
| *** exploreshaifali has quit IRC | 10:17 | |
| *** ekarlso- has quit IRC | 10:24 | |
| *** ekarlso- has joined #openstack-zaqar | 10:24 | |
| flwang1 | flaper87: ping | 10:36 |
| *** kgriffs has quit IRC | 10:51 | |
| *** kgriffs|afk has joined #openstack-zaqar | 10:51 | |
| *** kgriffs|afk is now known as kgriffs | 10:51 | |
| openstackgerrit | Fei Long Wang proposed openstack/zaqar: Implement mongodb driver for notifications https://review.openstack.org/139792 | 10:53 |
| *** cpallares has joined #openstack-zaqar | 11:09 | |
| cpallares | ping | 11:11 |
| cpallares | .. | 11:11 |
| *** vkmc has joined #openstack-zaqar | 11:15 | |
| vkmc | morning >:) | 11:18 |
| *** shaifali_ is now known as exploreshaifali | 11:23 | |
| exploreshaifali | good morning vkmc cpallares | 11:23 |
| cpallares | hello exploreshaifali, how are you doing? | 11:23 |
| cpallares | exploreshaifali: how was your new years? :) | 11:24 |
| exploreshaifali | cpallares, it was grate :D | 11:24 |
| cpallares | exploreshaifali: that's good to hear :) | 11:24 |
| exploreshaifali | lots of food, dance,games and what about you | 11:24 |
| vkmc | exploreshaifali, hey there!! | 11:24 |
| exploreshaifali | vkmc, when will flaper87 back | 11:25 |
| exploreshaifali | I am facing problem in queue_controller | 11:25 |
| exploreshaifali | :P | 11:25 |
| vkmc | I don't know really | 11:25 |
| vkmc | do you write to him or have regular meetings with him? | 11:26 |
| exploreshaifali | okay :) | 11:26 |
| exploreshaifali | vkmc, oh no, but that's a good idea | 11:26 |
| exploreshaifali | I will him for same :) | 11:26 |
| cpallares | exploreshaifali: I was wondering the same thing | 11:27 |
| flwang1 | :) | 11:27 |
| * cpallares sets out a gummy bear trap for flaper87 | 11:27 | |
| exploreshaifali | cpallares, lol | 11:27 |
| vkmc | heeeey flwang o/ | 11:28 |
| exploreshaifali | vkmc, I am trying debug in Zaqar, and just run one test case using testr yet getting error | 11:28 |
| exploreshaifali | pasted http://dpaste.com/0YBVH7N | 11:28 |
| exploreshaifali | flwang1, Happy New Year! | 11:29 |
| flwang1 | exploreshaifali: thanks, you too | 11:29 |
| exploreshaifali | :) | 11:29 |
| exploreshaifali | Duplicate test id is the problem | 11:30 |
| exploreshaifali | vkmc, please can you look at https://review.openstack.org/#/c/144340/ | 11:35 |
| vkmc | sure | 11:35 |
| vkmc | I'm checking out the gate, there is something wrong | 11:35 |
| exploreshaifali | vkmc, thanks, but I am sure this is not gate issue as I am getting same error in my local achine too | 11:36 |
| exploreshaifali | it is due to https://review.openstack.org/#/c/144340/2/zaqar/storage/mongodb/driver.py | 11:36 |
| vkmc | sure, what I meant is that I'll check it out as soon as I finish with the gate | 11:37 |
| exploreshaifali | line 208 | 11:37 |
| exploreshaifali | yeah, just to inform you getting *TypeError: 'QueueController' object is not callable* | 11:38 |
| exploreshaifali | due to line 208 in mongodb/driver module | 11:38 |
| exploreshaifali | I think I have done something wrong there | 11:39 |
| * cpallares changed her twitter handle to ᕕ( ᐛ )ᕗ | 12:05 | |
| ekarlso- | ello folks :p | 12:08 |
| ekarlso- | flaper87: about to chat riak ? | 12:08 |
| vkmc | ekarlso-, hey there! | 12:08 |
| vkmc | ekarlso-, would you like to write a poc in an external repo? | 12:09 |
| ekarlso- | vkmc: yeah, only problem I have is the object TTL stuff u guys use atm | 12:14 |
| ekarlso- | not sure if that's possible with riak | 12:14 |
| vkmc | I see | 12:16 |
| ekarlso- | hence the d0h | 12:16 |
| vkmc | maybe there is an alternative way to achieve the same behaviour in riak? | 12:17 |
| ekarlso- | not sure atm | 12:17 |
| *** bradjones has quit IRC | 12:51 | |
| *** bradjones has joined #openstack-zaqar | 12:53 | |
| *** exploreshaifali has quit IRC | 13:29 | |
| *** SairaKhanum has joined #openstack-zaqar | 13:36 | |
| *** kgriffs has quit IRC | 13:38 | |
| *** kgriffs|afk has joined #openstack-zaqar | 13:38 | |
| *** kgriffs|afk is now known as kgriffs | 13:38 | |
| *** shibanis has joined #openstack-zaqar | 13:48 | |
| *** pcaruana has joined #openstack-zaqar | 13:49 | |
| vkmc | kgriffs, o/ | 13:55 |
| *** exploreshaifali has joined #openstack-zaqar | 14:03 | |
| *** sriram has joined #openstack-zaqar | 14:06 | |
| SairaKhanum | hi all!! | 14:08 |
| vkmc | hey there SairaKhanum! | 14:09 |
| shibanis | hi | 14:09 |
| SairaKhanum | happy new year:-[ | 14:09 |
| shibanis | So we read a few things about zaqar and its architecture | 14:09 |
| SairaKhanum | now we would like know what excatly we are supposed to do.. | 14:10 |
| shibanis | Do we have to start with fixing bugs? | 14:12 |
| vkmc | that's cool | 14:15 |
| vkmc | sure, feel free to browse the list of opened bugs | 14:15 |
| vkmc | and take one | 14:16 |
| vkmc | or if you see something is not working as expected, report the bug | 14:16 |
| vkmc | or you can also contribute with reviewing | 14:16 |
| vkmc | doc writing | 14:16 |
| vkmc | testing | 14:16 |
| shibanis | okay | 14:17 |
| vkmc | you can also help with the client here https://launchpad.net/openstack/python-zaqarclient | 14:18 |
| vkmc | Zaqar bugs are listed here https://bugs.launchpad.net/openstack/zaqar | 14:18 |
| vkmc | Zaqar features, aka blueprints, are listed here https://blueprints.launchpad.net/zaqar | 14:19 |
| vkmc | same for the client | 14:19 |
| shibanis | okay thanks .. will go through them :) | 14:22 |
| SairaKhanum | small doubt.. | 14:25 |
| SairaKhanum | we can take up any proposal from blueprints right? | 14:25 |
| SairaKhanum | and who gives permission to implement... | 14:26 |
| *** amitgandhinz has joined #openstack-zaqar | 14:27 | |
| *** amitgandhinz has quit IRC | 14:27 | |
| exploreshaifali | I think that is the beauty of opensource, you don't need any ones permission, right vkmc | 14:27 |
| exploreshaifali | and moreover all code you will write will go through a review process | 14:28 |
| vkmc | exploreshaifali, exactly! I totally agree | 14:28 |
| vkmc | we have a roadmap | 14:28 |
| shibanis | so we can chose any blueprint and start designing | 14:28 |
| exploreshaifali | if other developers find that code god than only that code will be added to real software | 14:28 |
| shibanis | cool :) | 14:29 |
| vkmc | but you are encouraged to take a blueprint and discuss about it in the weekly meeting | 14:29 |
| shibanis | okay | 14:29 |
| vkmc | the roadmap only establish what we have to do... but its not a constrain | 14:29 |
| *** mpanetta has joined #openstack-zaqar | 14:29 | |
| exploreshaifali | vkmc, please can you help me in figure out what blunder I did | 14:31 |
| exploreshaifali | I am getting *TypeError: 'QueueController' object is not callable* | 14:31 |
| SairaKhanum | thank you guys .... | 14:32 |
| vkmc | exploreshaifali, sure, just a sec | 14:33 |
| vkmc | SairaKhanum, thanks to you! | 14:33 |
| *** amitgandhinz has joined #openstack-zaqar | 14:33 | |
| *** amitgandhinz has quit IRC | 14:33 | |
| *** amitgandhinz has joined #openstack-zaqar | 14:35 | |
| *** shibanis has left #openstack-zaqar | 14:36 | |
| *** SairaKhanum has left #openstack-zaqar | 14:36 | |
| *** kgriffs has quit IRC | 14:41 | |
| *** kgriffs|afk has joined #openstack-zaqar | 14:41 | |
| *** kgriffs|afk is now known as kgriffs | 14:42 | |
| *** ametts has joined #openstack-zaqar | 14:45 | |
| *** openstack has joined #openstack-zaqar | 15:06 | |
| *** kgriffs has quit IRC | 15:28 | |
| *** kgriffs has joined #openstack-zaqar | 15:28 | |
| *** achanda has joined #openstack-zaqar | 15:30 | |
| *** exploreshaifali has quit IRC | 15:54 | |
| openstackgerrit | Victoria Martínez de la Cruz proposed openstack/zaqar: Version discovery for root URI https://review.openstack.org/130094 | 15:56 |
| *** miqui has joined #openstack-zaqar | 16:06 | |
| *** achanda has quit IRC | 16:06 | |
| *** openstackgerrit has quit IRC | 16:06 | |
| *** openstackgerrit has joined #openstack-zaqar | 16:07 | |
| *** achanda has joined #openstack-zaqar | 16:07 | |
| *** achanda has quit IRC | 16:12 | |
| *** achanda has joined #openstack-zaqar | 16:16 | |
| *** SairaKhanum has joined #openstack-zaqar | 16:31 | |
| *** SairaKhanum has left #openstack-zaqar | 16:31 | |
| *** achanda has quit IRC | 16:32 | |
| *** amalagon has quit IRC | 16:38 | |
| *** amalagon has joined #openstack-zaqar | 16:38 | |
| *** amalagon has quit IRC | 16:43 | |
| *** reed has joined #openstack-zaqar | 16:48 | |
| *** openstackgerrit has quit IRC | 16:51 | |
| *** openstackgerrit has joined #openstack-zaqar | 16:51 | |
| *** kgriffs has quit IRC | 16:52 | |
| *** kgriffs|afk has joined #openstack-zaqar | 16:52 | |
| *** kgriffs|afk is now known as kgriffs | 16:52 | |
| *** openstackgerrit has quit IRC | 17:04 | |
| *** openstackgerrit has joined #openstack-zaqar | 17:04 | |
| *** amalagon has joined #openstack-zaqar | 17:06 | |
| *** achanda has joined #openstack-zaqar | 17:15 | |
| *** achanda has quit IRC | 17:28 | |
| *** achanda has joined #openstack-zaqar | 17:29 | |
| *** achanda has quit IRC | 17:34 | |
| *** kgriffs has quit IRC | 17:36 | |
| *** kgriffs|afk has joined #openstack-zaqar | 17:36 | |
| *** kgriffs|afk is now known as kgriffs | 17:36 | |
| *** exploreshaifali has joined #openstack-zaqar | 17:48 | |
| *** achanda has joined #openstack-zaqar | 17:50 | |
| *** achanda has quit IRC | 17:59 | |
| *** SairaKhanum has joined #openstack-zaqar | 17:59 | |
| *** SairaKhanum has left #openstack-zaqar | 17:59 | |
| *** achanda has joined #openstack-zaqar | 17:59 | |
| *** achanda has quit IRC | 18:04 | |
| *** SairaKhanum has joined #openstack-zaqar | 18:16 | |
| *** SairaKhanum has left #openstack-zaqar | 18:16 | |
| *** achanda has joined #openstack-zaqar | 18:26 | |
| *** kgriffs is now known as kgriffs|afk | 18:28 | |
| *** achanda has quit IRC | 18:31 | |
| *** echevemaster has joined #openstack-zaqar | 18:35 | |
| * vkmc lurks | 18:35 | |
| vkmc | kragniz, what are you doing? it make me nervous when you are quiet | 18:36 |
| vkmc | I know you are contriving something evil | 18:37 |
| kragniz | vkmc: sitting in an aircraft at the moment | 18:41 |
| * kragniz puts phone in flight mode | 18:43 | |
| vkmc | :x | 18:49 |
| vkmc | exploreshaifali, I'm reviewing your patchset | 18:50 |
| exploreshaifali | vkmc, thanks!!! | 18:52 |
| sriram | kragniz: if you are using gogo inflight wifi or something. http://www.techworm.net/2015/01/gogo-inflight-internet-intentionally-issuing-fake-ssl-certificates.html | 18:52 |
| *** amalagon has quit IRC | 19:00 | |
| *** amalagon has joined #openstack-zaqar | 19:00 | |
| vkmc | exploreshaifali, so... for what I understood | 19:06 |
| vkmc | you are having a problem with polimorphism | 19:06 |
| exploreshaifali | vkmc, okay but where | 19:06 |
| vkmc | polymorphism* | 19:07 |
| vkmc | your change in the driver base class is affecting the implementations of the interface and its tests | 19:08 |
| exploreshaifali | problem is due to https://github.com/openstack/zaqar/blob/master/zaqar/storage/base.py#L154 | 19:09 |
| exploreshaifali | as per what I understood | 19:09 |
| vkmc | so basically, when you instantiate the controldriver here https://review.openstack.org/#/c/144340/2/zaqar/storage/mongodb/driver.py | 19:09 |
| vkmc | you are instantiating a class with abstract methods | 19:10 |
| exploreshaifali | before instantiating abstract method is overriden | 19:11 |
| exploreshaifali | queue_controller() is overridden in both ControllDrive and also in DataDrive | 19:11 |
| vkmc | but its not in the correct context | 19:11 |
| vkmc | that's why its trying to instantiate the abstract one | 19:11 |
| vkmc | and that is why I said that your problem there is with polymorphism | 19:11 |
| exploreshaifali | sorry but what you mean by "its not a correct context" | 19:12 |
| exploreshaifali | vkmc, ^ | 19:14 |
| vkmc | I'm trying to think in a simple way to put it | 19:14 |
| vkmc | but the best thing you can do to notice what is going on | 19:15 |
| vkmc | is to put a breakpoint in one of the tests failing | 19:15 |
| vkmc | and inspect what is going on | 19:15 |
| exploreshaifali | vkmc, yeah here comes debugging | 19:15 |
| *** kgriffs|afk is now known as kgriffs | 19:16 | |
| vkmc | when you do this 'return ControlDriver(self.conf, self.cache).queue_controller()' | 19:16 |
| exploreshaifali | vkmc, I am running *python -m testtools.run discover --load-list tests/unit/storage/test_impl_mongodb.py* | 19:16 |
| vkmc | you are returning an instance of the queue_controller | 19:16 |
| exploreshaifali | and getting error http://dpaste.com/0YBVH7N | 19:16 |
| vkmc | while there is another instance already around being used | 19:16 |
| vkmc | those are two different objects | 19:17 |
| *** JAHoagie has joined #openstack-zaqar | 19:17 | |
| exploreshaifali | vkmc, and this is happening in base.py? | 19:17 |
| exploreshaifali | or in driver.py? | 19:17 |
| vkmc | its happening when you start the server | 19:17 |
| exploreshaifali | sorry not getting | 19:19 |
| exploreshaifali | lets first solve the debugging issue | 19:19 |
| exploreshaifali | look hre http://dpaste.com/0YBVH7N | 19:19 |
| vkmc | instead of looking at the output when trying to run the tests | 19:20 |
| vkmc | check out what happens when you try to start the server | 19:20 |
| vkmc | that way we can focus only on a set of errors | 19:20 |
| vkmc | I have it in verbose mode, sorry | 19:21 |
| vkmc | http://paste.openstack.org/show/155631/ | 19:21 |
| *** flwang1 has quit IRC | 19:22 | |
| vkmc | so... 'ControlDriver' object has no attribute 'queues_database' | 19:22 |
| vkmc | let's grep the code | 19:22 |
| vkmc | its in driver.py... for the storage you are using | 19:23 |
| vkmc | in my case, its mongodb/driver.py | 19:23 |
| exploreshaifali | vkmc, yes queues_database is not in ControlDriver | 19:24 |
| exploreshaifali | I figured it out | 19:24 |
| vkmc | ok so... first, let's make the server works | 19:24 |
| exploreshaifali | copied the code in ControlDriver | 19:24 |
| vkmc | and then we can keep debugging the tests | 19:25 |
| exploreshaifali | but still got same erro | 19:25 |
| exploreshaifali | *error | 19:25 |
| vkmc | same output? | 19:25 |
| exploreshaifali | not exactly same | 19:26 |
| exploreshaifali | but the error is same | 19:26 |
| exploreshaifali | that is | 19:26 |
| exploreshaifali | QueueController object is not callable | 19:26 |
| exploreshaifali | queues_databse error is not there now | 19:26 |
| exploreshaifali | but QueueController one is still there | 19:27 |
| exploreshaifali | vkmc, I don't understand about the server stuff you are saying | 19:29 |
| exploreshaifali | first make the server work means what? | 19:29 |
| vkmc | exploreshaifali, start the server simply | 19:29 |
| exploreshaifali | okay :) | 19:29 |
| vkmc | what is the output of 'zaqar-server'? | 19:29 |
| vkmc | the best way to debug something is to run and see if it works | 19:30 |
| exploreshaifali | vkmc, okay that is a tip for me ;) | 19:30 |
| vkmc | running the whole test suite may make things more difficult at early stages of development | 19:30 |
| exploreshaifali | ok, so here after running the server getting same *'QueueController' object is not callable* error | 19:31 |
| vkmc | paste the full trace | 19:32 |
| exploreshaifali | vkmc, http://paste.openstack.org/show/155633/ but avoid that 999999 and @@@@ stuff | 19:34 |
| exploreshaifali | I added that while debugging manually | 19:34 |
| vkmc | :O | 19:34 |
| vkmc | put something funnier | 19:35 |
| vkmc | like \o)\o/(o/ | 19:35 |
| exploreshaifali | yeah :D I should have | 19:35 |
| vkmc | ok so | 19:35 |
| exploreshaifali | lol | 19:35 |
| vkmc | haha | 19:35 |
| exploreshaifali | so now what next | 19:36 |
| vkmc | that is what I'm trying to figure out | 19:37 |
| exploreshaifali | okay!! | 19:40 |
| exploreshaifali | btw vkmc will be mentoring for GSoC? :P | 19:41 |
| vkmc | haha dunno, it would be fun | 19:42 |
| exploreshaifali | yeah loooots of, for any one who will be working under you :) | 19:42 |
| vkmc | with* | 19:42 |
| vkmc | :D | 19:42 |
| exploreshaifali | yes yes... with ;) | 19:43 |
| exploreshaifali | vkmc, I have one more problem | 19:47 |
| exploreshaifali | without making any change in code when I tried to debug with testr | 19:47 |
| exploreshaifali | get error http://dpaste.com/0YBVH7N | 19:47 |
| vkmc | ok, I got distracted debugging your code | 19:54 |
| vkmc | so | 19:54 |
| vkmc | I'm stil hitting errors, but I think it's a matter of finding all the dependencies and fixing them | 19:54 |
| vkmc | once you can get your server up | 19:55 |
| vkmc | then you run the tests | 19:55 |
| vkmc | but for now... we have to make sure that the migration works | 19:55 |
| vkmc | try to make the server up | 19:59 |
| vkmc | and if you cannot fix it, then I'll keep debugging | 19:59 |
| vkmc | let me see what happens with the tests | 19:59 |
| *** flwang1 has joined #openstack-zaqar | 20:00 | |
| vkmc | exploreshaifali, ok... and the tests... I'm able to reproduce | 20:08 |
| vkmc | so its a bug | 20:08 |
| vkmc | feel free to report it | 20:09 |
| exploreshaifali | :D | 20:09 |
| exploreshaifali | vkmc, but I don't know about the problem there | 20:09 |
| vkmc | good catch :) | 20:09 |
| vkmc | how to fix it you mean? | 20:09 |
| exploreshaifali | yes | 20:09 |
| vkmc | so probably its something caused by the recent renaming | 20:10 |
| vkmc | we removed the queues directory | 20:10 |
| vkmc | that's my first hunch | 20:10 |
| vkmc | so we have to check where is the tests name overlap and rename one | 20:10 |
| exploreshaifali | yup.... yup | 20:11 |
| exploreshaifali | I will just file the bug for now | 20:11 |
| exploreshaifali | and let see if we can make to fix it | 20:11 |
| exploreshaifali | I facing that problem since 31-Dec :/ and was pulling my hairs why am I unable to debug zaqar | 20:12 |
| exploreshaifali | loool :D | 20:13 |
| vkmc | daaaamn, thats bad to hear | 20:15 |
| exploreshaifali | but now I laugh at myself..... It was fun:D | 20:25 |
| kragniz | sriram: nah, the plan hadn't taken off and I had 3g | 20:25 |
| kragniz | s/plan/plane/ | 20:26 |
| vkmc | kragniz, where are yaaaa | 20:28 |
| kragniz | nowhere exciting! | 20:30 |
| kragniz | traveling cambridge -> ireland | 20:31 |
| kragniz | now eating jelly babies in the airport waiting for my coach to arrive | 20:32 |
| * kragniz leads a wild life | 20:32 | |
| vkmc | cool :D | 20:34 |
| *** kgriffs is now known as kgriffs|afk | 20:35 | |
| openstackgerrit | Fei Long Wang proposed openstack/zaqar: Implement mongodb driver for notifications https://review.openstack.org/139792 | 20:42 |
| openstackgerrit | Fei Long Wang proposed openstack/zaqar: Add REST API for notification/subscriptions https://review.openstack.org/142432 | 20:42 |
| sriram | kragniz: aah cool | 20:42 |
| vkmc | sriram, maybe you hit into this before... running python -m testtools.run discover fails with a duplicateid error | 20:44 |
| vkmc | any idea why it could be? | 20:44 |
| vkmc | there are no duplicateids | 20:44 |
| sriram | hmm, nope vkmc have not run into it. :/ | 20:45 |
| vkmc | hmm k k | 20:45 |
| vkmc | flwang, ^? | 20:45 |
| * sriram is embarrased to admit that its been long since he ran zaqar's tests. | 20:45 | |
| vkmc | haha no problem :) | 20:46 |
| vkmc | I usually run tox... so I have no clue why this is happening with testr | 20:46 |
| sriram | you should also be able to run nose right? | 20:47 |
| sriram | I've had the version test fail before, I just had to re-create the environment. | 20:49 |
| sriram | but that doesnt to be related in this case. | 20:50 |
| vkmc | lets see | 20:50 |
| *** nakul_cpani has joined #openstack-zaqar | 20:51 | |
| nakul_cpani | Hi guys !! | 20:54 |
| vkmc | hey nakul_cpani! | 20:54 |
| nakul_cpani | hi vkmc | 20:54 |
| nakul_cpani | hope u had good new year vacation | 20:55 |
| nakul_cpani | :) | 20:55 |
| vkmc | it was good thanks, and yours? | 20:55 |
| nakul_cpani | was quite good :) | 20:55 |
| vkmc | sriram, nose works :| | 20:55 |
| vkmc | nakul_cpani, good to hear | 20:56 |
| nakul_cpani | where do you guys work ? | 20:56 |
| flwang | vkmc: what's up? | 20:57 |
| nakul_cpani | i mean place | 20:57 |
| vkmc | flwang, running python -m testtools.run discover fails with a duplicateid error, any clue why this is happening? | 20:57 |
| vkmc | flwang, there are no duplicateids | 20:57 |
| *** kgriffs|afk is now known as kgriffs | 20:58 | |
| vkmc | exploreshaifali, https://bugs.launchpad.net/zaqar/+bug/1407786 | 20:59 |
| flwang | vkmc: hmm... any specific reason why we have to use that command? | 20:59 |
| vkmc | flwang, to be able to use pdb | 21:00 |
| flwang | ah, ok, to use pdb in unit test, is it? | 21:00 |
| vkmc | https://wiki.openstack.org/wiki/Testr#Debugging_.28pdb.29_Tests | 21:00 |
| vkmc | exactly | 21:00 |
| sriram | vkmc: huh, intereesting! | 21:01 |
| vkmc | flwang, hi btw :) | 21:03 |
| sriram | but using pdb shouldnt matter, no matter which test runner you are using right? | 21:03 |
| sriram | I might be wrong. | 21:03 |
| sriram | just wondering | 21:03 |
| vkmc | I guess so | 21:12 |
| vkmc | I just put a breakpoint and run nose | 21:13 |
| vkmc | and died there | 21:13 |
| vkmc | the halting problem all over again | 21:13 |
| flwang | vkmc: I didn't see the duplicate id issue but some other error like 'xxx object has no attribute url_prefix' | 21:13 |
| vkmc | flwang, which storage backend? | 21:14 |
| exploreshaifali | I haven't put any break point and tried to run testr command *python -m testtools.run discover --load-list tests/unit/storage/test_pool_queues.py* | 21:16 |
| exploreshaifali | then got duplicate test id issue | 21:16 |
| vkmc | exploreshaifali, add that to launchpad :) | 21:19 |
| exploreshaifali | yeah vkmc about to do that :P | 21:19 |
| exploreshaifali | thanks :) | 21:20 |
| vkmc | exploreshaifali, np | 21:20 |
| JAHoagie | when is the next Zaqar meeting? | 21:24 |
| vkmc | hi JAHoagie! | 21:24 |
| vkmc | JAHoagie, next week | 21:24 |
| JAHoagie | hi vkmc! | 21:24 |
| JAHoagie | vkmc, next Monday at 15:00? | 21:26 |
| vkmc | JAHoagie, yes | 21:26 |
| vkmc | JAHoagie, if there is something urgent you want to discuss, feel free to ping us here | 21:26 |
| JAHoagie | was mostly going to "listen" in | 21:27 |
| *** fifieldt_ has quit IRC | 21:27 | |
| vkmc | sure :) | 21:27 |
| *** fifieldt_ has joined #openstack-zaqar | 21:28 | |
| JAHoagie | that meeting is during my morning commute time, but I'll try to join | 21:28 |
| sriram | vkmc: did you set the breakpoint using nose's set_trace? | 21:28 |
| sriram | instead of import pdb; pdb.set_trace() | 21:29 |
| vkmc | JAHoagie, no problem... we always upload the logs in https://wiki.openstack.org/wiki/Meetings/Zaqar | 21:29 |
| vkmc | sriram, no :T | 21:29 |
| * vkmc not used to nose | 21:29 | |
| flwang | vkmc: sqlite as the management plane and mongo as the data plane | 21:30 |
| sriram | vkmc: its easy from nose import set_trace; set_trace() | 21:30 |
| vkmc | sriram, thanks | 21:30 |
| sriram | oops | 21:30 |
| sriram | from nose.tools | 21:30 |
| vkmc | :D | 21:31 |
| sriram | lemme know if it helps! :) | 21:31 |
| vkmc | sure thing :D | 21:32 |
| exploreshaifali | I tried to use nose.tools but it also didn't worked | 21:40 |
| exploreshaifali | vkmc, https://bugs.launchpad.net/zaqar/+bug/1407786 | 21:40 |
| exploreshaifali | 2 comments added | 21:40 |
| vkmc | thanks exploreshaifali | 21:40 |
| * exploreshaifali feeling shy...... all time when vkmc says thanks :P | 21:41 | |
| vkmc | haha naah | 21:41 |
| vkmc | you shouldn't | 21:41 |
| sriram | hmm, then it isnt the stdout issue I haved faced before.. | 21:41 |
| exploreshaifali | :) | 21:41 |
| exploreshaifali | if you guyz are also able to generate same Error then I think it could be due to the one that vkmc pointed | 21:43 |
| exploreshaifali | recently queue was removed | 21:43 |
| exploreshaifali | so could had renamed something | 21:43 |
| vkmc | nope, I just verified | 21:44 |
| vkmc | is not that | 21:44 |
| exploreshaifali | okay :) | 21:44 |
| vkmc | for some reason when running the test its taking the _ off | 21:44 |
| vkmc | or truncating it | 21:44 |
| vkmc | so the test for v1_0 and v1_1 is the same | 21:44 |
| sriram | is it some special character? | 21:44 |
| vkmc | '_' | 21:45 |
| sriram | yeah, is that character a special char to testr or something | 21:45 |
| vkmc | that I'm not sure | 21:46 |
| *** sriram has quit IRC | 22:04 | |
| *** nakul_cpani has left #openstack-zaqar | 22:21 | |
| vkmc | kgriffs, where are you, we miss you | 22:21 |
| * cpallares looks for kgriffs | 22:23 | |
| cpallares | kgriffs: knock knock | 22:23 |
| kgriffs | hi | 22:23 |
| vkmc | :D :D :D | 22:23 |
| kgriffs | man, I miss all of you! | 22:23 |
| cpallares | kgriffs: :D | 22:23 |
| vkmc | come procrastinate with us | 22:23 |
| cpallares | haha | 22:23 |
| vkmc | how have been doing kgriffs? merry xmas and happy new year :) | 22:24 |
| kgriffs | I've been reassigned to some other internal work at Rackspace and haven't had the chance to come up for air the past few weeks. :( | 22:24 |
| kgriffs | vkmc: yes, happy holidays! | 22:24 |
| cpallares | kgriffs: Happy Holidays and New Years!!!!!!!!!! | 22:24 |
| cpallares | kgriffs: Are you still working on barbican? | 22:24 |
| cpallares | kgriffs: You should just join Red Hat :) | 22:25 |
| kgriffs | got to play in the snow with the kids last week in Colorado. Part of me misses the snow, but part of me (the part that would have to shovel it) is glad to live in Austin, TX | 22:25 |
| vkmc | hahaha | 22:25 |
| kgriffs | cpallares: lol. I see how it is. :) | 22:25 |
| flwang | kgriffs: ping | 22:26 |
| kgriffs | o/ | 22:26 |
| cpallares | kgriffs: You secretly want to! You pass as a red hatter during summits :P | 22:26 |
| flwang | kgriffs: happy new year! | 22:26 |
| kgriffs | likewise, thanks! | 22:26 |
| flwang | kgriffs: and i have a question about the design of notifications | 22:26 |
| kgriffs | cpallares: heh | 22:26 |
| kgriffs | flwang: sure | 22:26 |
| flwang | kgriffs: for the supporting of pool | 22:26 |
| flwang | kgriffs: you know, for know, our current design of pooling is really depending on 'queue' | 22:27 |
| flwang | so for messages and claim, their rest api are look like '/queues/fizbit/messages' | 22:28 |
| flwang | but for notifications/subscriptions, we prefer to make it independent with queue | 22:29 |
| flwang | so the rest api will be like 'v2.0/subscriptions' | 22:29 |
| flwang | that means there is no queue name in the url | 22:29 |
| flwang | however, we will miss the 'queue' info to support pool | 22:30 |
| flwang | am I clear? | 22:30 |
| flwang | please let me know if I missed something | 22:30 |
| *** kgriffs has quit IRC | 22:33 | |
| *** exploreshaifali has quit IRC | 22:34 | |
| *** kgriffs has joined #openstack-zaqar | 22:36 | |
| kgriffs | hmm | 22:36 |
| kgriffs | flwang: let me see if I understand | 22:36 |
| kgriffs | say a user wants to subscribe to all messages posted to queue AKA topic "fizbit" | 22:37 |
| kgriffs | so, the user needs to tell me what topic they care about, and how they want to be notified | 22:38 |
| kgriffs | right? | 22:38 |
| flwang | that's ok | 22:38 |
| flwang | my point is because the subscriptions number maybe huge | 22:39 |
| flwang | so it would be nice if we can support 'pool' for subscriptions | 22:39 |
| flwang | however, it will make some troubles for the rest api design | 22:40 |
| flwang | because our current pool design is tight coupling with 'queue' | 22:41 |
| *** ametts has quit IRC | 22:57 | |
| *** ametts has joined #openstack-zaqar | 23:01 | |
| * kgriffs sorry, had to step away for a few minutes again | 23:06 | |
| *** bradjones has quit IRC | 23:06 | |
| kgriffs | so, let me think out loud for a minute | 23:07 |
| kgriffs | when creating a subscription, you would include in the post JSON document the topic you want "fizbit" and the notification, "webhook" (for example) | 23:08 |
| kgriffs | so you are worried that there may be a huge number of subscription lines that have to be looked up for each message? | 23:08 |
| * kgriffs has to go jump on a bus. will be back online in a few minutes | 23:09 | |
| *** bradjones has joined #openstack-zaqar | 23:10 | |
| *** mpanetta has quit IRC | 23:18 | |
| flwang | kgriffs: for creation, it's fine. but for get list, get, update, it would be hard to get the subscription if there is no queue info | 23:21 |
| kgriffs | oic | 23:22 |
| kgriffs | hmmm | 23:22 |
| kgriffs | let's break it down case by case | 23:23 |
| kgriffs | first, list subscriptions | 23:23 |
| kgriffs | I may just want to show all subscriptions, say if I only have a few and want to display in Horizon or something | 23:23 |
| kgriffs | Then I would just GET /v2.0/subscriptions | 23:24 |
| kgriffs | but if I want to find all subscriptions for a particular topic AKA queue | 23:24 |
| kgriffs | I could do: | 23:24 |
| flwang | ok | 23:24 |
| flwang | subscripts?topic=xxx | 23:24 |
| kgriffs | right, that's what I was about to type | 23:24 |
| flwang | subscriptions?source=xxx | 23:24 |
| flwang | yep, it should be ok | 23:25 |
| flwang | but how about the get and update? | 23:25 |
| kgriffs | I think you need a subscription ID for that. The ID would be returned after creating the subscription, but also when listing | 23:25 |
| flwang | yep, user will have the id | 23:25 |
| kgriffs | We could specify that the ID could be any string, but in reality we'd probably implement as UUID | 23:26 |
| kgriffs | so in that case you would say | 23:26 |
| kgriffs | GET subscriptions/9C416F21-4032-4489-92AC-5C69AC0D6B53 | 23:26 |
| kgriffs | PATCH or PUT subscriptions/9C416F21-4032-4489-92AC-5C69AC0D6B53 | 23:27 |
| kgriffs | something like that | 23:27 |
| flwang | now we're using oid of mongo | 23:27 |
| kgriffs | IMO in the API docs we should say that the client should just use the ID or URI verbatim - the ID could be anything | 23:28 |
| kgriffs | so, it could be OID or whatever we want but the client shouldn't care | 23:28 |
| flwang | yep | 23:28 |
| flwang | so the case turns to be, we have the id passed in, then how to get the correct subscription record with it? | 23:29 |
| flwang | given there maybe many pools | 23:29 |
| flwang | with different database types | 23:29 |
| kgriffs | hmm | 23:29 |
| kgriffs | there is a database that is used for the pool catalog, right? | 23:30 |
| kgriffs | we could put subscriptions there - it becomes the "control plane" database. | 23:30 |
| kgriffs | so you only have one place to look. | 23:30 |
| kgriffs | subscriptions should be very cacheable (won't change very often), so you could easily offload lookups from that database | 23:31 |
| flwang | kgriffs: yep, i see. but current pool catalog is depending on 'queue' | 23:31 |
| flwang | without queue, what's the 'key' for routing the request? | 23:31 |
| *** amitgandhinz has quit IRC | 23:33 | |
| kgriffs | hmm, I was thinking that there would be a separate table / collection for the subscriptions | 23:33 |
| flwang | kgriffs: sorry, i have to run and will be back in 40 mins | 23:33 |
| kgriffs | kk | 23:33 |
| kgriffs | I will keep typing for when you return | 23:33 |
| kgriffs | :) | 23:33 |
| flwang | kgriffs: haha, awesome :) | 23:33 |
| kgriffs | in that subscription collection you would have a list of records | 23:33 |
| kgriffs | each contains | 23:33 |
| kgriffs | ID | Topic | Project ID | Notification Type | Notification Settings | 23:36 |
| kgriffs | Notification Settings is a blob (MessagePack or JSON) with the schema dependent on Notification Type | 23:37 |
| kgriffs | So, when you want to CRUD subscriptions, you are simply interacting with this collection, using subscription IDs as the primary key | 23:39 |
| kgriffs | for the data plane, when a message comes in, zaqar would first persist the message, then send a "notify subscribers" task to a worker pool (however that is implemented) | 23:41 |
| kgriffs | one of the workers would then take that task and look up all the subscribers (if any) for the given (topic, project_id) | 23:42 |
| kgriffs | it would then schedule one ore more new tasks that other workers would take to actually deliver the notifications. Alternatively, you could have the web request handler do the lookup for subscriptions and cut out the middle man, but you probably want web requests to return as quickly as possible. something to think about, anyway... | 23:44 |
| kgriffs | plenty of details to work out around exactly the most efficient / scalable way to farm out the notification tasks, but you get the idea... | 23:45 |
| *** exploreshaifali has joined #openstack-zaqar | 23:49 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!