| *** wanghao has joined #openstack-zaqar | 00:39 | |
| *** shu-mutou_ has joined #openstack-zaqar | 00:41 | |
| *** openstackgerrit has joined #openstack-zaqar | 00:48 | |
| *** wanghao_ has joined #openstack-zaqar | 01:04 | |
| *** wanghao has quit IRC | 01:07 | |
| *** wanghao_ has quit IRC | 01:08 | |
| *** wanghao has joined #openstack-zaqar | 01:09 | |
| *** gecong has joined #openstack-zaqar | 01:09 | |
| flwang | wanghao: gecong: ping | 01:50 |
|---|---|---|
| flwang | wxy: ping | 01:50 |
| flwang | could we get this in https://review.openstack.org/529205 to remove the block for the gate? | 01:50 |
| wanghao | flwang: here | 01:51 |
| wanghao | flwang: sure, will review it later. | 01:51 |
| flwang | thanks | 01:52 |
| gecong | hi | 01:54 |
| flwang | btw, we had some interesting discussion tc channel | 01:55 |
| gecong | could we review https://review.openstack.org/#/c/528996/ | 01:55 |
| flwang | here is the log if you guys interested in | 01:55 |
| flwang | http://eavesdrop.openstack.org/irclogs/%23openstack-tc/%23openstack-tc.2017-12-20.log.html#t2017-12-20T21:46:11 | 01:55 |
| flwang | gecong: we can merge https://review.openstack.org/#/c/528996/ but i prefer to trying the test skip with the separated patch i'm proposing | 01:56 |
| gecong | ok | 01:57 |
| openstackgerrit | gecong proposed openstack/zaqar master: Replace pymongo deprecated api https://review.openstack.org/528996 | 02:00 |
| *** harlowja has quit IRC | 02:07 | |
| *** wanghao_ has joined #openstack-zaqar | 02:16 | |
| *** wanghao has quit IRC | 02:20 | |
| flwang | the idea in that discussion is, can zaqar create a new service to listen infra messaging bus and then forward them into tenant user's queues | 02:22 |
| wxy | flwang: Of cause | 02:49 |
| wxy | flwang: But should we do that? | 02:50 |
| flwang | wxy: i'm not sure, but what's i'm thinking is we can create a separated service which is a contrib of zaqar | 02:51 |
| flwang | like vpnaas of neutron | 02:51 |
| wxy | flwang: it's a little out of Zaqar's scope | 02:52 |
| wxy | flwang: how about creating a new kind of queue, this queue can listen some area. | 02:53 |
| flwang | wxy: you can think out of zaqar | 02:53 |
| flwang | i'm trying to address a problem in openstack which is highly related to zaqar | 02:53 |
| flwang | wxy: that way is hacking zaqar totally to meet a special requirement, right? | 02:54 |
| flwang | in other words, i'm thinking if we can create another tiny service which is a good friend with zaqar | 02:54 |
| flwang | who can help address the user-facing notifications issue | 02:55 |
| flwang | but it's not a part of zaqar | 02:55 |
| flwang | could be a separated repo | 02:55 |
| flwang | could be based on therve's work https://github.com/therve/osn/tree/master/osn | 02:55 |
| wxy | flwang: thinking, I'll read the log again. | 02:56 |
| flwang | wxy: thanks | 02:57 |
| yangzhenyu | flwang, | 03:09 |
| yangzhenyu | In fact, this demand is real. We have done before a message notification service, this service is to monitor the notification queue of rabbitmq, and then sent through the Restful api to our other manage service. Of course, we did filtering to send only messages of interest. | 03:09 |
| gecong | Ok I will read the log carefully, thanks | 03:10 |
| yangzhenyu | flwang, At present, many practices are non-stop query the status of resources to update in real time. This is very system resources and can not be real-time. For example, the internal shutdown of the virtual machine, the management platform does not know now is off, this time need to be able to send this message to the cloud platform. | 03:13 |
| flwang | yangzhenyu: i see. but that's out of the goal | 03:15 |
| yangzhenyu | flwang, The message is actually available in the message queue of rabbitmq, but no way to custom sent to other places. | 03:15 |
| yangzhenyu | flwang, I think almost. | 03:16 |
| *** wanghao_ has quit IRC | 03:18 | |
| *** wanghao has joined #openstack-zaqar | 03:18 | |
| flwang | yangzhenyu: yep, that's the problem we're trying to resolve | 03:18 |
| yangzhenyu | flwang, Just need a way to sent the tenant user resource notification message out, can be sent to the user, can also be sent to other service platforms. | 03:18 |
| flwang | yes | 03:19 |
| yangzhenyu | listen and send | 03:19 |
| yangzhenyu | I think it is valuable. | 03:21 |
| flwang | yangzhenyu: https://github.com/therve/osn/tree/master/osn take a look at this | 03:22 |
| flwang | that's the thing we are talking about | 03:23 |
| flwang | maybe we can create a new tiny service (as a good friend with zaqar) to do the job | 03:23 |
| yangzhenyu | yes, such as this repo, add a zaqar driver and some api or zaqar use ws to interact. I think it works | 03:32 |
| wxy | flwang: If i remember correctly, we had a discussion about service queue in Zaqar which tried to integrate with SearchLight, right? | 03:51 |
| wxy | flwang: So what's the main reason that now we need a new component to collect notification? | 03:53 |
| flwang | wxy: i think one of the reasons is searchlight is not active anymore | 03:53 |
| flwang | we can just wait, don't do anything | 03:54 |
| flwang | because we don't have to | 03:54 |
| flwang | i mention this just because i think it's good thing to have in openstack world | 03:54 |
| wxy | flwang: Is it possible that adding a new RESTful driver in oslo.messaging? | 03:58 |
| flwang | wxy: that's another option which has been mentioned many times | 03:59 |
| wxy | flwang: yeah. because it's the easiest way. :) | 03:59 |
| flwang | but it's not suitable for zaqar's position | 04:02 |
| wxy | flwang: I'm afraid that creating a new component seems to reinvent the wheel. If no community support, it'll be just a toy. | 04:06 |
| flwang | i see | 04:09 |
| *** flwang1 has quit IRC | 04:11 | |
| openstackgerrit | Merged openstack/zaqar master: Skip period test for message https://review.openstack.org/529205 | 04:34 |
| openstackgerrit | gecong proposed openstack/zaqar master: Replace pymongo deprecated api https://review.openstack.org/528996 | 05:06 |
| openstackgerrit | gecong proposed openstack/zaqar master: Support delayed queues for mongo https://review.openstack.org/508669 | 05:09 |
| openstackgerrit | gecong proposed openstack/zaqar master: Updated from global requirements https://review.openstack.org/528936 | 05:12 |
| *** shu-mutou_ has quit IRC | 05:13 | |
| *** shu-mutou_ has joined #openstack-zaqar | 05:13 | |
| openstackgerrit | gecong proposed openstack/zaqar master: Convert zaqar-tox-integration to native Zuul v3 https://review.openstack.org/528509 | 05:14 |
| openstackgerrit | gecong proposed openstack/zaqar master: Test delay queues https://review.openstack.org/513283 | 05:16 |
| *** gengchc2 has joined #openstack-zaqar | 05:18 | |
| gecong | @gengchc2 http://eavesdrop.openstack.org/irclogs/%23openstack-tc/%23openstack-tc.2017-12-20.log.html#t2017-12-20T21:46:11 | 05:19 |
| gecong | @gengchc2 https://github.com/therve/osn/tree/master/osn take a look at this | 05:19 |
| *** flwang1 has joined #openstack-zaqar | 05:23 | |
| *** flwang1 has quit IRC | 05:25 | |
| *** flwang1 has joined #openstack-zaqar | 05:48 | |
| *** vkmc has quit IRC | 06:26 | |
| *** robcresswell has quit IRC | 06:26 | |
| *** ameade has quit IRC | 06:26 | |
| *** vkmc has joined #openstack-zaqar | 06:31 | |
| *** robcresswell has joined #openstack-zaqar | 06:31 | |
| *** ameade has joined #openstack-zaqar | 06:31 | |
| *** vkmc has quit IRC | 06:48 | |
| *** robcresswell has quit IRC | 06:48 | |
| *** ameade has quit IRC | 06:48 | |
| wanghao | flwang: after reading the log, I prefer to propose a small middleware to support get notificaiton from mb to users's queue. | 06:51 |
| wanghao | flwang: that will not only be used with openstack projects, it may could work with other communtity, like k8s. | 06:53 |
| wanghao | wxy: so adding Restful driver in oslo.messaging is a option, but the scope of use cases may be smaller. | 06:54 |
| *** robcresswell has joined #openstack-zaqar | 07:25 | |
| *** ameade has joined #openstack-zaqar | 07:25 | |
| *** vkmc has joined #openstack-zaqar | 07:30 | |
| *** vkmc has quit IRC | 07:30 | |
| *** vkmc has joined #openstack-zaqar | 07:30 | |
| *** gecong has quit IRC | 07:34 | |
| *** rcernin has quit IRC | 07:37 | |
| *** flwang1 has quit IRC | 07:42 | |
| *** flwang1 has joined #openstack-zaqar | 07:51 | |
| *** pcaruana has joined #openstack-zaqar | 07:58 | |
| *** AJaeger has joined #openstack-zaqar | 08:18 | |
| AJaeger | zaqar cores, could you review https://review.openstack.org/528509 to improve your CI jobs, please? | 08:18 |
| AJaeger | flwang1: ^ | 08:26 |
| flwang1 | AJaeger: no problem, thanks | 08:54 |
| AJaeger | thanks | 08:59 |
| therve | flwang1, https://github.com/therve/nabu was supposed to be the next version of osn :) | 09:01 |
| wanghao | AJaeger: sure, got it. | 09:04 |
| wanghao | therve: thanks, that's great | 09:04 |
| therve | I don't know, but it's something for sure :) | 09:06 |
| wanghao | flwang1: BTW, post my comments above, prefer to adding a small thing to work with Zaqar. | 09:07 |
| wanghao | therve: yep | 09:07 |
| therve | But that is basically aodh+event alarms though | 09:08 |
| wanghao | we can also support to push notifications from rabbitmq base on nabu. | 09:12 |
| wanghao | this is a good start I think. | 09:12 |
| *** shu-mutou_ has quit IRC | 09:19 | |
| *** wanghao has quit IRC | 09:34 | |
| *** wanghao_ has joined #openstack-zaqar | 09:34 | |
| AJaeger | thanks, therve | 09:40 |
| therve | No problem! | 09:41 |
| therve | Thank you | 09:41 |
| *** jtomasek has joined #openstack-zaqar | 09:42 | |
| flwang1 | therve: you mean nabu is based on aodh+event alarm? | 09:42 |
| therve | flwang1, No that's basically what it does | 09:43 |
| therve | aodh+event alarm can do what nabu does, I think | 09:43 |
| flwang1 | therve: yep, i see | 09:43 |
| flwang1 | therve: did you see the long discussion we (zaneb, Matt R and I ) had in tc channel? | 09:44 |
| therve | I had a quick look | 09:44 |
| flwang1 | what's your opinion? | 09:45 |
| flwang1 | do you think it's creating a new wheel if we gonna add a new tiny friend for zaqar | 09:45 |
| therve | A new wheel? | 09:47 |
| therve | I don't know, I somewhat lost interest in that problem | 09:49 |
| *** gengchc2 has quit IRC | 09:51 | |
| openstackgerrit | Merged openstack/zaqar master: Support delayed queues for mongo https://review.openstack.org/508669 | 09:54 |
| *** jtomasek has quit IRC | 10:07 | |
| flwang1 | therve: ok, i see. | 10:07 |
| flwang1 | thanks for the input | 10:07 |
| *** shu-mutou_HOME has joined #openstack-zaqar | 10:10 | |
| *** wanghao has joined #openstack-zaqar | 10:32 | |
| *** wanghao_ has quit IRC | 10:36 | |
| *** pcaruana has quit IRC | 10:43 | |
| *** wanghao has quit IRC | 10:45 | |
| *** wanghao has joined #openstack-zaqar | 10:45 | |
| *** wanghao_ has joined #openstack-zaqar | 11:00 | |
| *** wanghao has quit IRC | 11:01 | |
| therve | flwang1, I hadn't thought about the nova metadata | 11:14 |
| therve | That's an interesting way to solve the issue, if those are indeed part of the notification payload | 11:15 |
| *** shu-mutou_HOME has quit IRC | 11:17 | |
| openstackgerrit | Merged openstack/zaqar master: Replace pymongo deprecated api https://review.openstack.org/528996 | 11:20 |
| openstackgerrit | Merged openstack/zaqar master: Convert zaqar-tox-integration to native Zuul v3 https://review.openstack.org/528509 | 11:28 |
| *** AJaeger has left #openstack-zaqar | 11:35 | |
| flwang1 | therve: but that way still needs changes in nova | 11:56 |
| *** wanghao_ has quit IRC | 12:30 | |
| *** wanghao has joined #openstack-zaqar | 12:30 | |
| *** wanghao has quit IRC | 12:31 | |
| *** wanghao has joined #openstack-zaqar | 12:31 | |
| *** wanghao has quit IRC | 12:31 | |
| *** wanghao has joined #openstack-zaqar | 12:32 | |
| *** wanghao has quit IRC | 12:32 | |
| *** wanghao has joined #openstack-zaqar | 12:33 | |
| *** wanghao has quit IRC | 12:33 | |
| *** wanghao has joined #openstack-zaqar | 12:33 | |
| *** wanghao has quit IRC | 12:34 | |
| *** wanghao has joined #openstack-zaqar | 12:34 | |
| *** wanghao has quit IRC | 12:35 | |
| *** wanghao has joined #openstack-zaqar | 12:35 | |
| *** wanghao has quit IRC | 12:35 | |
| *** wanghao has joined #openstack-zaqar | 12:36 | |
| *** wanghao has quit IRC | 12:36 | |
| *** wanghao has joined #openstack-zaqar | 12:36 | |
| *** wanghao has quit IRC | 12:37 | |
| *** wanghao has joined #openstack-zaqar | 12:38 | |
| *** wanghao has quit IRC | 12:42 | |
| therve | flwang1, Not the way I see it no | 12:49 |
| therve | You "just" need a notification sink, which does stuff if those metadata are present | 12:50 |
| therve | It needs something like service queues though | 12:50 |
| flwang1 | no change needed in nova ? that's great | 12:52 |
| *** jtomasek has joined #openstack-zaqar | 13:34 | |
| *** jtomasek has quit IRC | 14:37 | |
| *** pcaruana has joined #openstack-zaqar | 17:04 | |
| openstackgerrit | Merged openstack/zaqar master: Doc of delayed queues https://review.openstack.org/513284 | 17:14 |
| openstackgerrit | Merged openstack/zaqar master: Test delay queues https://review.openstack.org/513283 | 17:30 |
| *** harlowja has joined #openstack-zaqar | 18:16 | |
| *** flwang1 has quit IRC | 19:29 | |
| *** pcaruana has quit IRC | 20:12 | |
| *** pcaruana has joined #openstack-zaqar | 20:29 | |
| *** flwang1 has joined #openstack-zaqar | 20:37 | |
| *** pcaruana has quit IRC | 20:42 | |
| *** flwang has quit IRC | 21:23 | |
| *** flwang has joined #openstack-zaqar | 21:36 | |
| *** andreaf has quit IRC | 21:55 | |
| *** andreaf has joined #openstack-zaqar | 21:56 | |
| *** rcernin has joined #openstack-zaqar | 22:24 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!