Tuesday, 2015-01-20

*** achanda has quit IRC00:02
*** achanda has joined #openstack-zaqar00:02
*** achanda has quit IRC00:03
*** achanda has joined #openstack-zaqar00:04
flwangvkmc: still around?00:26
*** achanda has quit IRC00:30
openstackgerritFei Long Wang proposed openstack/zaqar: Check marker before accessing it  https://review.openstack.org/14839601:01
vkmcflwang, sure01:12
flwangvkmc: https://review.openstack.org/#/c/148396/01:17
flwangvkmc: I can't find a good way to test it, so leave it as it is01:17
vkmcflwang, oh cool :)01:19
* vkmc clicks01:19
*** echevemaster has joined #openstack-zaqar01:42
*** echevemaster has quit IRC01:42
*** kgriffs is now known as kgriffs|afk01:51
*** cpallares has quit IRC02:14
openstackgerritMATSUMURA, Shuntaro proposed openstack/python-zaqarclient: Fix request by claim message  https://review.openstack.org/14841002:42
*** achanda has joined #openstack-zaqar03:20
openstackgerritVictoria Martínez de la Cruz proposed openstack/zaqar: Update json schema for API v1_1  https://review.openstack.org/14841403:27
*** achanda has quit IRC03:41
*** achanda has joined #openstack-zaqar03:44
*** achanda has quit IRC03:51
*** achanda has joined #openstack-zaqar04:02
*** flwang1 has quit IRC04:08
*** cpallares has joined #openstack-zaqar04:22
*** cpallares has quit IRC04:23
*** csoukup has joined #openstack-zaqar05:09
*** achanda has quit IRC05:09
*** achanda has joined #openstack-zaqar05:11
*** csoukup has quit IRC05:36
*** davideagnello has joined #openstack-zaqar06:48
*** davideagnello has quit IRC06:52
*** miqui has quit IRC06:54
*** achanda has quit IRC07:15
*** miqui_ has quit IRC07:54
*** exploreshaifali has joined #openstack-zaqar08:04
*** dynarro has joined #openstack-zaqar08:12
openstackgerritFlavio Percoco proposed openstack/zaqar: Split MongoDB's default implementation from FIFO's  https://review.openstack.org/14829708:26
exploreshaifalimorning!! :)08:27
flaper87exploreshaifali: I'm not here08:27
flaper87:P08:27
flaper87exploreshaifali: goooood morning08:27
exploreshaifalihahaha08:28
exploreshaifaliflaper87, I can see that :P08:28
openstackgerritShaifali Agrawal proposed openstack/zaqar: Removing QueueController from data to control plane  https://review.openstack.org/14434008:28
exploreshaifaliflaper87, please give an eye to ^08:31
flaper87exploreshaifali: will do08:31
flaper87exploreshaifali: commented with some nits08:34
exploreshaifaliflaper87, okay. Thanks, I have made changes only for mongodb atm08:35
flaper87exploreshaifali: mind addressing those tiny nits asap so we can review on a clean patch set08:38
flaper87?08:38
exploreshaifaliflaper87, yes, just looking on them :)08:39
*** yfujioka has joined #openstack-zaqar08:40
*** davideagnello has joined #openstack-zaqar08:41
*** davideagnello has quit IRC08:45
exploreshaifaliflaper87, we need control_driver instance only for driver.DataDriver class08:47
exploreshaifaliright08:47
exploreshaifalibut if we bring `self.control_driver = control_driver` inside base class, than08:47
exploreshaifaliother DataDriver class such as pooling.DataDriver and pipeline.DataDriver will also affect08:48
exploreshaifaliflaper87, does these two also need instance of control_driver?08:48
flaper87exploreshaifali: All subclasses of DataDriver need control_driver08:48
flaper87we need to keep the constructor consistent08:49
openstackgerritMerged openstack/zaqar: Check marker before accessing it  https://review.openstack.org/14839608:49
exploreshaifaliflaper87, constructor consistent part is okay, but why all subclasses of DataDrive need control_drive?08:50
exploreshaifaliflaper87, we need control_drive just for passing QueueController to DataDriver08:50
exploreshaifalithat is done only in driver.DataDriver classes08:50
exploreshaifaliokay, sorry08:51
exploreshaifalipipeline.DataDriver also have queue_controller08:51
flaper87:)08:51
openstackgerritShaifali Agrawal proposed openstack/zaqar: Removing QueueController from data to control plane  https://review.openstack.org/14434009:17
openstackgerritShaifali Agrawal proposed openstack/zaqar: Removing QueueController from data to control plane  https://review.openstack.org/14434009:25
exploreshaifaliflaper87, ^09:25
exploreshaifaliI am getting errors pasted here : http://dpaste.com/1WHNNKZ09:35
* flaper87 clicks09:41
flaper87exploreshaifali: why are you passing the control_driver to the ControlDriver constructor?09:44
flaper87exploreshaifali: https://review.openstack.org/#/c/144340/8/zaqar/storage/base.py,cm09:44
flaper87you just need it in the DataDriver09:45
flaper87wait, wrong link. This one: https://review.openstack.org/#/c/144340/8/zaqar/storage/sqlalchemy/driver.py,cm09:45
flaper87also, pls, move the `queue_controller` definition to the base class09:46
exploreshaifalicontrol_driver is passes inside ConrolDriver constructor too because https://github.com/openstack/zaqar/blob/master/zaqar/storage/utils.py#L90-L93 here09:47
exploreshaifaliin invoke_args09:47
exploreshaifaliwe will pass control_drive too along with cache and cong09:47
exploreshaifali*conf09:48
flaper87exploreshaifali: yup but you need to change that function, not the constructors :P09:48
flaper87As in, if control_driver is passed to that function, then you add it to the invoke_args09:48
exploreshaifaliflaper87, okay, i can make if - else statement09:48
flaper87otherwise, you don't09:48
exploreshaifalifor that09:48
flaper87exploreshaifali: perfect09:48
*** flwang1 has joined #openstack-zaqar09:57
flwang1flaper87: ping09:57
flaper87flwang1: pong09:58
flwang1flaper87: do you have a few minutes to discus the notification service?09:58
flaper87flwang1: yes10:03
flwang1so given the subscription work is almost done, so it's time to take a look the notification sending part10:03
flaper87flwang1: sure10:06
flwang1so the first question is do we need a separate worker for notification sending?10:06
flwang1or just call taskflow in message posting to start a new task/job10:08
openstackgerritShaifali Agrawal proposed openstack/zaqar: Removing QueueController from data to control plane  https://review.openstack.org/14434010:20
flaper87flwang1: by worker you mean a separte zaqar instance running ?10:24
flaper87I'd like to avoid that for now, TBH.10:24
flwang1flaper87: ok10:25
flwang1so does that mean we will have to inject some code into message posting?10:25
flaper87flwang1: what do you mean?10:31
flwang1flaper87: i'm talking about where to trigger the notification sending?10:31
exploreshaifaliflaper87, if possible please review the patch, list of error is same http://dpaste.com/1WHNNKZ10:34
flaper87flwang1: in zaqar's web head, no?10:37
flaper87mmh, not sure I follow your question, flwang1  :(10:37
*** dynarro has quit IRC10:37
flwang1flaper87: hmm... sorry for the bad expression10:37
flwang1based on my understanding, the basic work flow of notification service will be like:10:38
flwang11. create subscription10:38
flwang12. post message on the queue10:38
flwang13. check if there is subscription for the queue10:39
flwang14. sending notification to the subscriber10:39
flwang1now the step 1 is almost done10:39
flwang1and we already implemented the step 210:39
flwang1so i'm taking about the step 3 and 410:39
flwang1and the question is mainly about10:41
flwang1'where'10:41
flwang1where to hack the code10:41
flaper87ah mmh, ok10:42
flaper87sorry, I'm slow today10:42
flaper87This is what I had in mind: We've a pipeline that we should probably use10:42
flaper87That'd be: Post the message and then call (in a coroutine) the code for sending the notifications10:43
flaper87https://git.openstack.org/cgit/openstack/zaqar/tree/zaqar/common/pipeline.py10:43
flaper87that's the pipeline I'm talking about10:43
flaper87I think we should *always* have a pipeline, regardless we're running in pooling mode or not10:44
flaper87thoughts ?10:44
flwang1yep, that's what i mentioned with you before10:44
flwang1flaper87: but if we want to leverage https://git.openstack.org/cgit/openstack/zaqar/tree/zaqar/common/pipeline.py for message posting, we may need refactor current message posting code10:45
flwang1is that ok for you?10:45
flwang1i mean refactor the code at here https://git.openstack.org/cgit/openstack/zaqar/tree/zaqar/transport/wsgi/v2_0/messages.py#n153 to support pipeline10:47
flaper87flwang1: I was thinking we would inject the pipeline here:https://git.openstack.org/cgit/openstack/zaqar/tree/zaqar/bootstrap.py#n9610:48
flwang1flaper87: hmm...10:53
flwang1seems we're talking about different levels10:53
flwang1so how can you leverage the pipeline for notification sending given currently the pipeline is only for data driver level, but i think the 'sending' is an operation/menthod level10:54
flaper87flwang1: well, the pipeline is a set of pipelined objects that have some methods implemented10:56
flaper87flwang1: I think we could define a "post_message" method in the 'SubscriberDriver' that will be called everytime a message is posted10:56
flaper87the difference is that the `post_message` method in the SubscriberDriver will trigger the publishing of such message(s) under the assumption that they were all published10:57
flaper87(otherwise it'd have failed earlier in the pipeline)10:57
flaper87does that make any sense?10:58
* flwang1 is thinking...11:01
* flaper87 sends some thinking energies to flwang111:05
flaper87flwang1: gtg, lunch.11:05
flaper87bbib11:05
*** yfujioka has quit IRC11:14
vkmcmorning :)11:33
kragnizvkmc: morning!11:33
vkmckragniz, hiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii11:34
exploreshaifalimorning vkmc kragniz  :)11:41
vkmcexploreshaifali, hiiiiiiiiiiiiiiiiiiiiiiiiiiiii to you too :D11:42
exploreshaifali:D11:43
* vkmc hopes nobody else comes... too much energy put in two first hi's 11:43
exploreshaifalilol11:43
exploreshaifaliwake up every one11:43
vkmcshhh exploreshaifali haha11:44
exploreshaifali:D11:45
*** hiii_vkmc has joined #openstack-zaqar11:45
hiii_vkmchiiiiiiiiiiiiiiiii11:45
exploreshaifalihiii_vkmc, hiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii11:46
exploreshaifali:P11:46
*** hiii_vkmc has quit IRC11:46
exploreshaifalibtw vkmc need your thoughts on http://dpaste.com/1WHNNKZ11:46
flaper87flwang1: ?11:46
flwang1flaper87: yep?11:46
flaper87flwang1: your last status was: "thinkink"11:46
flaper87I'm polling11:47
flaper87:P11:47
flwang1flaper87: yep, i need to write some code to check if it's workable11:47
flaper87flwang1: ok ok11:47
flaper87flwang1: let me know11:47
flaper87thanks11:47
exploreshaifalivkmc, error is occuring due to https://github.com/openstack/zaqar/blob/master/zaqar/storage/utils.py#L9911:48
flaper87exploreshaifali: https://review.openstack.org/#/c/144340/9/zaqar/storage/sqlalchemy/driver.py,cm11:48
flaper87the control driver does not accept a control_driver instance11:49
* vkmc clicks11:49
*** dynarro has joined #openstack-zaqar11:49
flaper87exploreshaifali: you are getting that error because stevedore fails to load the sqlalchemy driver11:49
vkmcflaper87, hola chico :)11:49
vkmcexploreshaifali, what flaper87 said ^11:49
flaper87vkmc: hola caracola! Como estás?11:49
flaper87exploreshaifali: what vkmc said that I said ^11:50
* flaper87 starts an infinite loop11:50
vkmcexploreshaifali, what flaper87 said that I said he said11:50
exploreshaifaliOMG.......but I know how to use break statement :P11:50
exploreshaifali:D11:50
flaper87exploreshaifali: commented on the patch11:51
vkmcflaper87, bien bien, leyendo el backlog :) y tu?11:51
exploreshaifaliflaper87, ok, Thanks!11:51
flaper87vkmc: https://review.openstack.org/#/c/147912/ >.>11:52
flaper87merge conflicts, great! #NOT11:52
vkmcwhaaaaaaaaaat? c'mon11:53
flaper87vkmc: when you get a chance, could you please review my patches?11:59
flaper87moving forward with the capabilities thing11:59
vkmcflaper87, of course! :)12:00
exploreshaifaliflaper87, *re-add this line* for base.py under subscription_controller.py?12:01
exploreshaifalihttps://review.openstack.org/#/c/144340/9/zaqar/storage/base.py12:01
flwang1flaper87: vkmc: https://review.openstack.org/#/c/142432/12:01
flwang1pls help review it ^12:02
flwang1thanks :)12:02
flwang1offline now12:02
flaper87exploreshaifali: yes... According to pep8, there should be 2 lines in between global definitions12:02
flaper87flwang1: have a great night12:02
exploreshaifaliok!12:02
vkmcflwang, good night! ttyt :)12:05
vkmcflwang1, ^12:05
flaper87vkmc: ssssssshhhhhhhhhhhhhhhh, flwang1 flwang1 flwang1 flwang1  is SLEEPING12:07
vkmcflaper87, btw https://review.openstack.org/#/c/148414/12:10
vkmcthat is a WIP12:10
vkmcbut any feedback is welcome12:10
*** pcaruana has joined #openstack-zaqar12:17
*** davideagnello has joined #openstack-zaqar12:18
*** davideagnello has quit IRC12:23
openstackgerritShaifali Agrawal proposed openstack/zaqar: Removing QueueController from data to control plane  https://review.openstack.org/14434012:35
exploreshaifaliflaper87, vkmc still getting same error, after removing control_driver parameter form ControlDriver class of sqlalchemy12:49
exploreshaifalidriver12:49
exploreshaifalisee the latest patch12:49
*** jchai has joined #openstack-zaqar13:00
flaper87exploreshaifali: commented13:06
*** exploreshaifali has quit IRC13:08
*** exploreshaifali has joined #openstack-zaqar13:11
*** exploreshaifali has quit IRC13:31
openstackgerritShaifali Agrawal proposed openstack/zaqar: Removing QueueController from data to control plane  https://review.openstack.org/14434013:33
vkmcmahna mahna13:44
openstackgerritLouis Taylor proposed openstack/zaqar: Move oslo libraries out of the oslo namespace  https://review.openstack.org/14856213:46
openstackgerritFlavio Percoco proposed openstack/zaqar: Use pool's capabilities for flavors in v2  https://review.openstack.org/14856613:53
* flaper87 is happy to have finally found the time to get this thing done ^13:53
*** sriram has joined #openstack-zaqar14:11
*** flwang1 has quit IRC14:11
*** csoukup has joined #openstack-zaqar14:11
*** mpanetta has joined #openstack-zaqar14:28
*** dynarro_ has joined #openstack-zaqar14:40
*** dynarro has quit IRC14:40
*** amitgandhinz has joined #openstack-zaqar14:41
*** cpallares has joined #openstack-zaqar15:07
*** jasondotstar has joined #openstack-zaqar15:16
*** kgriffs|afk is now known as kgriffs15:18
*** dynarro has joined #openstack-zaqar15:21
*** dynarro_ has quit IRC15:21
*** diga has quit IRC15:43
*** diga has joined #openstack-zaqar15:44
*** achanda has joined #openstack-zaqar15:50
*** achanda has quit IRC15:55
*** diga has quit IRC15:59
*** amalagon has quit IRC16:15
*** dynarro has quit IRC16:18
*** amalagon has joined #openstack-zaqar16:34
*** achanda has joined #openstack-zaqar16:58
*** sriram has quit IRC17:03
*** achanda has quit IRC17:07
*** jchai is now known as jchai_afk17:12
*** jchai_afk is now known as jchai17:14
*** davideagnello has joined #openstack-zaqar17:15
*** sriram has joined #openstack-zaqar17:39
jasondotstarhttps://wiki.openstack.org/wiki/Zaqar/Give_Zaqar_a_try#Set_up_a_simple_Zaqar_deployment  <= is this pretty current?17:40
jasondotstarmorning, btw.17:40
kragnizjasondotstar: it looks pretty up to date (tox -e genconfig is my benchmark)17:42
jasondotstarkragniz, thx. wanna see if it works17:47
*** davideagnello has quit IRC17:57
*** pcaruana has quit IRC18:00
*** csoukup has quit IRC18:08
*** achanda has joined #openstack-zaqar18:11
jasondotstarflaper87, going to use this as a basis for starting the puppet code...18:21
*** achanda has quit IRC18:28
*** jchai is now known as jchai_afk18:38
*** achanda has joined #openstack-zaqar18:50
*** davideagnello has joined #openstack-zaqar19:06
*** davideagnello has quit IRC19:07
*** jchai_afk is now known as jchai19:10
*** achanda has quit IRC19:29
*** flwang1 has joined #openstack-zaqar19:30
*** flwang1 has quit IRC19:32
*** achanda has joined #openstack-zaqar19:32
*** exploreshaifali has joined #openstack-zaqar19:33
*** openstackstatus has joined #openstack-zaqar19:46
*** ChanServ sets mode: +v openstackstatus19:46
-openstackstatus- NOTICE: Gerrit restarted to address likely memory leak leading to server slowness. Sorry if you were caught in the restart19:54
*** achanda has quit IRC20:03
flaper87jasondotstar: hey hey20:13
flaper87jasondotstar: so, yeah. That should be pretty much up-to-date20:14
flaper87jasondotstar: except for [drivers:storage:mongodb]20:14
flaper87jasondotstar: which is now [drivers:message_store:mongodb]20:15
*** exploreshaifali has quit IRC20:17
*** flwang1 has joined #openstack-zaqar20:18
*** achanda has joined #openstack-zaqar20:44
jasondotstarflaper87, ack.20:50
*** achanda has quit IRC20:51
vkmcjasondotstar, that is updated! :)20:54
vkmcthanks for working on it20:54
vkmcI'm mentally lagged :|21:02
kragnizvkmc: present for you! https://review.openstack.org/#/c/148562/21:07
vkmcI have a present for you too kragniz21:12
kragnizwhat can it be!?21:13
openstackgerritVictoria Martínez de la Cruz proposed openstack/zaqar: API handler and API v1.1 Queue endpoints  https://review.openstack.org/14128021:13
openstackgerritVictoria Martínez de la Cruz proposed openstack/zaqar: Adds message processing to WebSockets driver  https://review.openstack.org/14480321:13
openstackgerritVictoria Martínez de la Cruz proposed openstack/zaqar: Update json schema for API v1_1  https://review.openstack.org/14841421:13
openstackgerritVictoria Martínez de la Cruz proposed openstack/zaqar: Adds websockets driver to Zaqar server  https://review.openstack.org/14015521:13
openstackgerritVictoria Martínez de la Cruz proposed openstack/zaqar: Refactor Api classes to enhance consistency  https://review.openstack.org/14826521:13
vkmcthere21:13
kragnizeep21:13
vkmc:D21:14
vkmcahahahaha21:14
vkmcI'm kidding, that is for flaper8721:14
kragnizexcellent21:15
flaper87>.>21:17
vkmcbtw... how should I do to avoid gerrit to change topic names when creating dependencies?21:18
vkmcits bothering21:18
flaper87vkmc: use -t21:19
flaper87vkmc: git review -t your-topic21:20
vkmcmmh21:21
vkmcflaper87, but I have patchsets at the top with a different topic21:22
vkmc-t set the same topic for all changes21:22
*** davideagnello has joined #openstack-zaqar21:25
*** davideagnello has quit IRC21:29
*** davideagnello has joined #openstack-zaqar21:30
openstackgerritVictoria Martínez de la Cruz proposed openstack/zaqar: API handler and API v1.1 Queue endpoints  https://review.openstack.org/14128021:43
openstackgerritVictoria Martínez de la Cruz proposed openstack/zaqar: Refactor Api classes to enhance consistency  https://review.openstack.org/14826521:44
*** amitgandhinz has quit IRC21:44
*** amitgandhinz has joined #openstack-zaqar21:49
*** achanda has joined #openstack-zaqar21:57
*** davideagnello has quit IRC22:02
*** davideagnello has joined #openstack-zaqar22:04
*** achanda has quit IRC22:19
*** achanda has joined #openstack-zaqar22:20
*** mpanetta has quit IRC22:48
*** amitgandhinz has quit IRC22:50
*** sriram has quit IRC22:51
*** jchai has quit IRC22:54
*** davideagnello has quit IRC23:01
*** davideagnello has joined #openstack-zaqar23:02
*** achanda has quit IRC23:21
*** achanda has joined #openstack-zaqar23:21
openstackgerritVictoria Martínez de la Cruz proposed openstack/zaqar: API handler and API v1.1 Queue endpoints  https://review.openstack.org/14128023:45
openstackgerritVictoria Martínez de la Cruz proposed openstack/zaqar: Adds websockets driver to Zaqar server  https://review.openstack.org/14015523:46
openstackgerritVictoria Martínez de la Cruz proposed openstack/zaqar: Update json schema for API v1_1  https://review.openstack.org/14841423:47
openstackgerritVictoria Martínez de la Cruz proposed openstack/zaqar: Adds message processing to WebSockets driver  https://review.openstack.org/14480323:48
*** achanda has quit IRC23:48
*** achanda has joined #openstack-zaqar23:50

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