| *** JAHoagie has joined #openstack-zaqar | 00:03 | |
| *** achanda has quit IRC | 00:03 | |
| *** flwang has quit IRC | 00:03 | |
| *** achanda_ has quit IRC | 00:04 | |
| *** flwang has joined #openstack-zaqar | 00:04 | |
| *** flwang1 has quit IRC | 00:05 | |
| *** Qiming has quit IRC | 00:19 | |
| *** JAHoagie has quit IRC | 00:48 | |
| *** Qiming has joined #openstack-zaqar | 00:57 | |
| *** Qiming has quit IRC | 01:05 | |
| *** Qiming has joined #openstack-zaqar | 01:52 | |
| *** mdnadeem_ has joined #openstack-zaqar | 01:53 | |
| *** achanda has joined #openstack-zaqar | 02:33 | |
| *** david-ly_ has joined #openstack-zaqar | 02:39 | |
| *** david-lyle has quit IRC | 02:40 | |
| *** mdnadeem_ has quit IRC | 02:40 | |
| *** mdnadeem_ has joined #openstack-zaqar | 02:53 | |
| *** csoukup has joined #openstack-zaqar | 03:01 | |
| *** itisha has joined #openstack-zaqar | 03:04 | |
| *** mdnadeem_ has quit IRC | 03:12 | |
| *** david-ly_ is now known as david-lyle | 03:17 | |
| *** mdnadeem_ has joined #openstack-zaqar | 03:24 | |
| *** mdnadeem_ has quit IRC | 03:31 | |
| openstackgerrit | Fei Long Wang proposed openstack/zaqar: Add note for unreliable in README.rst https://review.openstack.org/226154 | 04:07 |
|---|---|---|
| *** JAHoagie has joined #openstack-zaqar | 04:37 | |
| *** JAHoagie has quit IRC | 04:40 | |
| *** stanchan has joined #openstack-zaqar | 04:42 | |
| *** csoukup has quit IRC | 04:47 | |
| *** JAHoagie has joined #openstack-zaqar | 04:51 | |
| *** diga has joined #openstack-zaqar | 05:02 | |
| *** mdnadeem_ has joined #openstack-zaqar | 05:09 | |
| *** mdnadeem_ has quit IRC | 06:33 | |
| *** itisha has quit IRC | 06:33 | |
| *** itisha has joined #openstack-zaqar | 06:36 | |
| *** stanchan has quit IRC | 06:48 | |
| *** david-lyle has quit IRC | 06:58 | |
| *** stanchan has joined #openstack-zaqar | 06:59 | |
| *** david-lyle has joined #openstack-zaqar | 07:14 | |
| *** achanda has quit IRC | 07:17 | |
| *** itisha has quit IRC | 07:41 | |
| *** boris-42 has quit IRC | 07:41 | |
| *** sebasmagri has quit IRC | 07:41 | |
| *** wanghao has quit IRC | 07:41 | |
| *** zhiyan has quit IRC | 07:41 | |
| *** ameade has quit IRC | 07:41 | |
| *** itisha has joined #openstack-zaqar | 08:14 | |
| *** mdnadeem_ has joined #openstack-zaqar | 08:29 | |
| *** mdnadeem_ has quit IRC | 08:52 | |
| *** mdnadeem_ has joined #openstack-zaqar | 08:56 | |
| *** Nesze has joined #openstack-zaqar | 09:04 | |
| *** mdnadeem_ has quit IRC | 09:05 | |
| openstackgerrit | Merged openstack/zaqar: Fix 500 error when adding pool to wrong group https://review.openstack.org/221122 | 09:06 |
| *** mdnadeem_ has joined #openstack-zaqar | 09:17 | |
| *** penchal has joined #openstack-zaqar | 09:19 | |
| penchal | Hi | 09:20 |
| penchal | i am trying to setup zaqar with redis backend and getting some issue | 09:21 |
| *** diga has quit IRC | 09:25 | |
| openstackgerrit | Flavio Percoco proposed openstack/zaqar: Update falcon requirements https://review.openstack.org/225269 | 09:26 |
| *** boris-42 has joined #openstack-zaqar | 09:26 | |
| Nesze | Hi, can I list more mongo uris in the zaqar’s config, like ‘uri = mongodb://<mongo1>, mongodb://<mongo2>, mongodb://<mongo3>’? | 09:39 |
| flaper87 | Nesze: nope, what you can do, though, is enable pooling and have different mongodb pools. A mongodb instance (URI) would be a pool | 09:40 |
| flaper87 | Nesze: zaqar will balance the load at a topic level | 09:40 |
| flaper87 | Nesze: unfortunately, I don't have a link pointing to docs about this | 09:41 |
| flaper87 | :( | 09:41 |
| penchal | can i use redis as a backend for zaqar ? | 09:50 |
| penchal | instead of mongo db | 09:50 |
| Nesze | thanks flaper87, so that would mean, that when I add a separate pool, pointing to another mongo, but these two mongos are in a cluster, and replication is enabled, I’d still have the same data in all the mongos. | 09:50 |
| *** ameade has joined #openstack-zaqar | 09:53 | |
| flaper87 | penchal: yes you can | 09:54 |
| *** zhiyan has joined #openstack-zaqar | 09:54 | |
| flaper87 | Nesze: ah, wait | 09:54 |
| flaper87 | I do not recommend doing that | 09:54 |
| flaper87 | If the reason you want to add multiple URLs is to provide the list of members in a replicaSet then yes, you can do that | 09:55 |
| flaper87 | However, I'd recommend you to add one and let pymongo figure the rest out | 09:55 |
| flaper87 | if you want to scale your data across multiple mongodb clusters, then you should use zaqar pools | 09:55 |
| *** wanghao has joined #openstack-zaqar | 09:56 | |
| *** stack_ has joined #openstack-zaqar | 09:56 | |
| *** stack_ has quit IRC | 09:56 | |
| *** sebasmagri has joined #openstack-zaqar | 09:56 | |
| penchal | getting some issue while configuring redis as a backend | 10:00 |
| Nesze | ta | 10:00 |
| penchal | 2015-09-22 09:05:57.271 29270 WARNING oslo_config.cfg [-] Option "verbose" from group "DEFAULT" is deprecated for removal. Its value may be silently ignored in the future. 2015-09-22 09:05:57.324 29270 WARNING oslo_config.cfg [-] Option "storage" from group "drivers" is deprecated. Use option "message_store" from group "drivers". 2015-09-22 09:06:27.597 29270 ERROR stevedore.extension [-] Could not load 'message_queue_handl | 10:00 |
| penchal | in zaqar.conf i used 6379 port for redis | 10:01 |
| penchal | uri = redis://127.0.0.1:6379 | 10:02 |
| *** flwang1 has joined #openstack-zaqar | 10:04 | |
| flwang1 | morning | 10:04 |
| *** Qiming has quit IRC | 10:08 | |
| *** Nesze has quit IRC | 10:09 | |
| *** itisha has quit IRC | 10:23 | |
| *** mdnadeem_ has quit IRC | 10:24 | |
| *** penchal has quit IRC | 10:42 | |
| *** itisha has joined #openstack-zaqar | 10:43 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/zaqar: Updated from global requirements https://review.openstack.org/226242 | 10:45 |
| *** Qiming has joined #openstack-zaqar | 10:56 | |
| openstackgerrit | Flavio Percoco proposed openstack/zaqar: Updated from global requirements https://review.openstack.org/226242 | 11:38 |
| *** Nesze has joined #openstack-zaqar | 11:44 | |
| *** flwang1 has quit IRC | 11:53 | |
| *** achanda has joined #openstack-zaqar | 12:41 | |
| *** achanda has quit IRC | 12:42 | |
| *** boris-42 has quit IRC | 12:49 | |
| flaper87 | therve: do you have 5 mins to take a look here? https://review.openstack.org/#/c/226242/ | 12:55 |
| therve | Ah cool there is only one failure | 12:59 |
| therve | I was wondering why https://review.openstack.org/#/c/225269/ was failing completely | 13:00 |
| *** sriram has joined #openstack-zaqar | 13:02 | |
| therve | flaper87, https://github.com/openstack/python-zaqarclient/blob/master/zaqarclient/tests/queues/queues.py#L365 ? | 13:03 |
| therve | Looks like the upgrade brings us some feature | 13:03 |
| openstackgerrit | Thomas Herve proposed openstack/python-zaqarclient: Fix functional test of get_many_messages https://review.openstack.org/226319 | 13:26 |
| openstackgerrit | Thomas Herve proposed openstack/python-zaqarclient: Fix functional test of get_many_messages https://review.openstack.org/226319 | 13:26 |
| therve | flaper87, ^^^ | 13:27 |
| *** Nesze has quit IRC | 13:28 | |
| *** Nesze has joined #openstack-zaqar | 13:29 | |
| vkmc | flaper87, if interested, I'm chatting with mrunge in -horizon | 13:30 |
| vkmc | re demo | 13:30 |
| flaper87 | vkmc: I am interested but I can't right now :( | 13:33 |
| flaper87 | therve: thanks. That seems to be the issue | 13:33 |
| openstackgerrit | Nate Johnston proposed openstack/zaqar: Use decorator instead of request logging in responders https://review.openstack.org/224882 | 13:33 |
| vkmc | flaper87, k, np | 13:33 |
| openstackgerrit | Flavio Percoco proposed openstack/zaqar: Updated from global requirements https://review.openstack.org/226242 | 13:37 |
| *** amitgandhinz has joined #openstack-zaqar | 13:51 | |
| vkmc | hey guys | 13:58 |
| vkmc | someone want to help mentoring someone for the Outreachy internship? | 13:58 |
| *** mpanetta has joined #openstack-zaqar | 13:58 | |
| vkmc | it doesn't take much time and will make a difference for Zaqar and for an intern professional career | 13:59 |
| vkmc | pleeeeeeeeeeeease | 13:59 |
| vkmc | I also have a project idea if you want | 13:59 |
| flaper87 | vkmc: I will if I'm not elected PTL for Glance | 14:00 |
| *** csoukup has joined #openstack-zaqar | 14:00 | |
| flaper87 | vkmc: more info to come on Thursday | 14:00 |
| flaper87 | :P | 14:00 |
| vkmc | flaper87, this Thursday finish the elections? nice | 14:01 |
| vkmc | crossing fingers for you | 14:01 |
| *** nate_gone is now known as njohnston | 14:08 | |
| flaper87 | therve: https://review.openstack.org/#/c/226319/ | 14:14 |
| *** exploreshaifali has joined #openstack-zaqar | 14:14 | |
| flaper87 | py26 doesn't support {} sets | 14:14 |
| openstackgerrit | Nate Johnston proposed openstack/zaqar: Use decorator instead of request logging in responders https://review.openstack.org/224882 | 14:18 |
| *** ametts has joined #openstack-zaqar | 14:19 | |
| Nesze | I’ve built zaqar from master, how complete is the v2 API implementation, is it almost done? Does it make sense to test my use cases with v2, or at this stage it’s better to stick with v1.1? | 14:27 |
| openstackgerrit | Thierry Carrez proposed openstack/zaqar: Open Mitaka development https://review.openstack.org/226347 | 14:28 |
| *** boris-42 has joined #openstack-zaqar | 14:30 | |
| Nesze | according to https://wiki.openstack.org/wiki/Zaqar/specs/api/2.0 is still in progress. Are there any other useful documentation sources, or I should start look into the code at last : ) ? | 14:30 |
| *** kgriffs|afk is now known as kgriffs | 14:35 | |
| *** exploreshaifali has quit IRC | 14:43 | |
| openstackgerrit | Thomas Herve proposed openstack/python-zaqarclient: Fix functional test of get_many_messages https://review.openstack.org/226319 | 14:49 |
| therve | screw python 2.6 :/ | 14:50 |
| *** Nesze has quit IRC | 15:07 | |
| njohnston | I was wondering if anyone could take a quick look at the failures for gate-tempest-dsvm-[redis|mongodb]-zaqar here; I have not seen these issues with any previous changesets, and I am trying to determine if suddenly my change is really failing those or if there is a transient gate issue or something. I just can't tell from the error logs. https://review.openstack.org/#/c/224882/ | 15:17 |
| vkmc | njohnston, checking it | 15:18 |
| njohnston | vkmc: Thanks! | 15:19 |
| *** mpanetta has quit IRC | 15:21 | |
| *** ametts has quit IRC | 15:25 | |
| jasondotstar | vkmc, flaper87, flwang: hey | 15:29 |
| jasondotstar | http://paste.openstack.org/show/473541/ | 15:30 |
| jasondotstar | installed zaqar w/ my puppet code (w00t!) | 15:30 |
| jasondotstar | but after installing mongodb (manually) I'm getting this error :-( | 15:30 |
| ryansb | \o/ what a champ | 15:30 |
| jasondotstar | digging down into the weeds now.... | 15:30 |
| jasondotstar | ryansb: thanks... little more to go before it's really really ready | 15:32 |
| ryansb | yeah, but that's a great step | 15:32 |
| *** Qiming has quit IRC | 15:36 | |
| vkmc | jasondotstar, wohoo | 15:44 |
| vkmc | let see that error | 15:44 |
| vkmc | jasondotstar, you need to add "unreliable=True" under [DEFAULT] in the zaqar.conf | 15:50 |
| *** stanchan has quit IRC | 15:54 | |
| *** whydidyoustealmy is now known as shakamunyi | 16:03 | |
| *** exploreshaifali has joined #openstack-zaqar | 16:12 | |
| *** JAHoagie has quit IRC | 16:27 | |
| exploreshaifali | how can I provide deprecation message for already existing commands? | 16:35 |
| jasondotstar | vkmc: ack | 16:38 |
| * jasondotstar tries | 16:39 | |
| *** mpanetta has joined #openstack-zaqar | 16:42 | |
| *** david-lyle has quit IRC | 16:46 | |
| *** achanda has joined #openstack-zaqar | 16:47 | |
| therve | Uh devstack doesn't respect requirements :/ | 16:50 |
| therve | njohnston, You need https://review.openstack.org/#/c/226242/ to land | 16:51 |
| therve | You can add a Depends-On | 16:51 |
| *** davideagnello has quit IRC | 17:08 | |
| jasondotstar | vkmc: how about this one: http://paste.openstack.org/show/473567/ | 17:08 |
| jasondotstar | stevedore ext issue? | 17:08 |
| *** davideagnello has joined #openstack-zaqar | 17:10 | |
| openstackgerrit | Nate Johnston proposed openstack/zaqar: Use decorator instead of request logging in responders https://review.openstack.org/224882 | 17:12 |
| njohnston | therve: Thanks much! | 17:12 |
| *** achanda_ has joined #openstack-zaqar | 17:17 | |
| *** achanda has quit IRC | 17:17 | |
| *** diga has joined #openstack-zaqar | 17:21 | |
| jasondotstar | vkmc: not sure what this 'WriteConcern' object has no attribute 'document' error is abt | 17:46 |
| *** exploreshaifali has quit IRC | 17:57 | |
| *** stanchan has joined #openstack-zaqar | 18:01 | |
| *** stanchan has quit IRC | 18:01 | |
| jasondotstar | flaper87, flwang ^ ne idea? | 18:08 |
| *** amitgandhinz has quit IRC | 18:10 | |
| *** kgriffs is now known as kgriffs|afk | 18:13 | |
| *** kgriffs|afk is now known as kgriffs | 18:13 | |
| vkmc | jasondotstar, could you paste me the configuration file? | 18:16 |
| jasondotstar | ya | 18:21 |
| jasondotstar | that's was going to be my next qn - a working sample from a zaqar install w/ a mongodb backend | 18:21 |
| jasondotstar | but let me get the one i have | 18:21 |
| *** achanda has joined #openstack-zaqar | 18:23 | |
| *** achanda_ has quit IRC | 18:24 | |
| *** diga has quit IRC | 18:27 | |
| *** achanda has quit IRC | 18:42 | |
| *** achanda has joined #openstack-zaqar | 18:48 | |
| jasondotstar | vkmc: http://paste.openstack.org/show/473590/ | 18:51 |
| *** ametts has joined #openstack-zaqar | 18:56 | |
| *** exploreshaifali has joined #openstack-zaqar | 19:14 | |
| *** flwang1 has joined #openstack-zaqar | 19:16 | |
| *** flwang1 has quit IRC | 19:25 | |
| *** achanda has quit IRC | 19:37 | |
| *** amitgandhinz has joined #openstack-zaqar | 19:40 | |
| *** bradjones has joined #openstack-zaqar | 19:46 | |
| *** bradjones has quit IRC | 19:46 | |
| *** bradjones has joined #openstack-zaqar | 19:46 | |
| *** itisha has quit IRC | 19:53 | |
| *** ametts has quit IRC | 20:02 | |
| *** csoukup has quit IRC | 20:07 | |
| *** JAHoagie has joined #openstack-zaqar | 20:13 | |
| *** achanda has joined #openstack-zaqar | 20:15 | |
| *** JAHoagie has quit IRC | 20:19 | |
| *** JAHoagie has joined #openstack-zaqar | 20:19 | |
| jasondotstar | vkmc: did you see the conf file yet? | 20:38 |
| vkmc | jasondotstar, I did, and got distracted with some other stuff :( sorry | 20:38 |
| vkmc | it looks fine | 20:38 |
| vkmc | its odd | 20:38 |
| vkmc | I thought it was outdated | 20:38 |
| jasondotstar | vkmc: no worries | 20:40 |
| jasondotstar | eventually, I'll have to dynamically generate the params in there but | 20:40 |
| jasondotstar | I'm just trying to make sure a basic one works | 20:40 |
| vkmc | yeah | 20:41 |
| jasondotstar | as i said, for kicks i have a mongodb installed as the backend | 20:41 |
| vkmc | its the first time I see that error actually | 20:41 |
| jasondotstar | hmmm ok | 20:41 |
| jasondotstar | glad it's not something goofy that i did. | 20:41 |
| vkmc | what version of Mongo are you using? | 20:41 |
| jasondotstar | ummm | 20:41 |
| jasondotstar | 1 sec | 20:41 |
| vkmc | oh no | 20:41 |
| jasondotstar | 2.6.10 ? | 20:42 |
| jasondotstar | vkmc: uh oh.... you said oh no | 20:42 |
| vkmc | haha | 20:42 |
| vkmc | its fine | 20:42 |
| vkmc | jasondotstar, comment out L39 | 20:42 |
| vkmc | storage=mongodb | 20:42 |
| jasondotstar | ah | 20:42 |
| jasondotstar | ok | 20:42 |
| jasondotstar | well, when i do that it's complaining about the sqlite driver | 20:45 |
| jasondotstar | is that the default? | 20:45 |
| vkmc | this is a minimum configuration file sample http://paste.openstack.org/show/473608/ | 20:45 |
| jasondotstar | vkmc: ^ | 20:45 |
| vkmc | no, the thing is that there are options that are deprecated | 20:45 |
| vkmc | and they are not properly documented | 20:46 |
| jasondotstar | oh ok | 20:46 |
| jasondotstar | i c | 20:46 |
| jasondotstar | ok | 20:46 |
| jasondotstar | perhaps i can help w/ that too (eventually) | 20:46 |
| jasondotstar | let me try it with your sample conf | 20:46 |
| vkmc | try with management_store = mongodb and message_store = mongodb for the storage driver selection | 20:46 |
| vkmc | sure | 20:46 |
| vkmc | I have websocket enabled | 20:46 |
| vkmc | so make sure you change that for wsgi | 20:47 |
| jasondotstar | oh | 20:50 |
| jasondotstar | ok | 20:50 |
| vkmc | transport=wsgi | 20:51 |
| vkmc | :) | 20:51 |
| jasondotstar | nope, I've got 1.8.1 | 20:54 |
| jasondotstar | vkmc: hmmmm. ran it in debug mode. saw this: http://paste.openstack.org/show/473613/ | 21:04 |
| jasondotstar | perhaps I'm not using a id/pass for mongo? | 21:04 |
| jasondotstar | bbiab | 21:05 |
| vkmc | yeah, you should use your credentials | 21:05 |
| vkmc | I'm not sure if the write_concern thing is related though | 21:05 |
| *** mpanetta has quit IRC | 21:05 | |
| *** sriram has quit IRC | 21:10 | |
| *** achanda has quit IRC | 21:21 | |
| *** JAHoagie has quit IRC | 22:18 | |
| *** angvp has quit IRC | 22:19 | |
| *** csoukup has joined #openstack-zaqar | 22:19 | |
| *** angvp has joined #openstack-zaqar | 22:20 | |
| *** angvp is now known as Guest87950 | 22:20 | |
| *** boris-42 has quit IRC | 22:22 | |
| *** achanda has joined #openstack-zaqar | 22:22 | |
| *** boris-42 has joined #openstack-zaqar | 22:24 | |
| *** achanda has quit IRC | 22:27 | |
| *** csoukup has quit IRC | 22:29 | |
| *** amitgandhinz has quit IRC | 22:30 | |
| *** Guest87950 is now known as angvp | 22:35 | |
| *** angvp has quit IRC | 22:35 | |
| *** angvp has joined #openstack-zaqar | 22:35 | |
| vkmc | flwang, good morning :) | 22:50 |
| *** kgriffs is now known as kgriffs|afk | 22:53 | |
| *** kgriffs|afk is now known as kgriffs | 22:54 | |
| jasondotstar | vkmc: yeah im not sure what that is either | 22:56 |
| vkmc | jasondotstar, did you try running it with your mongo creds? | 22:57 |
| vkmc | zaqar db that is | 22:57 |
| jasondotstar | flwang: morning | 22:57 |
| flwang | morning :) | 23:02 |
| vkmc | flwang, had a chat with mrunge earlier today | 23:04 |
| flwang | vkmc: anything i should be aware? | 23:06 |
| vkmc | flwang, not really, but I'll paste the log | 23:07 |
| vkmc | we only discussed a few how Horizon is currently managing the status update | 23:07 |
| vkmc | s | 23:07 |
| vkmc | and he mentioned they are mostly concerned about the tables updates | 23:07 |
| vkmc | the balloons notifications are not in their priorities | 23:08 |
| flwang | vkmc: right | 23:08 |
| vkmc | http://paste.openstack.org/show/473649/ | 23:08 |
| vkmc | full log | 23:08 |
| vkmc | no, wai | 23:09 |
| vkmc | t | 23:09 |
| vkmc | haha | 23:09 |
| *** flwang1 has joined #openstack-zaqar | 23:10 | |
| vkmc | http://paste.openstack.org/show/473650/ | 23:11 |
| vkmc | there, that's better | 23:11 |
| flwang | reading... | 23:12 |
| *** achanda has joined #openstack-zaqar | 23:12 | |
| vkmc | so I assume is the same you chatted with him | 23:22 |
| vkmc | for what I understand, they interact with the other services performing API calls directly | 23:23 |
| vkmc | https://github.com/openstack/horizon/tree/master/openstack_dashboard/api | 23:23 |
| vkmc | here they have all the clients | 23:23 |
| flwang | yep | 23:25 |
| flwang | i'm thinking the pros & cons compared with the patch mrunge mentioned | 23:26 |
| flwang | just had a quick review for the patch, i think we can just remove his self-cooked websocket https://review.openstack.org/#/c/40198/4/horizon-realtime.py and let the js https://review.openstack.org/#/c/40198/4/horizon/static/horizon/js/horizon.realtime.js listen to zaqar | 23:30 |
| vkmc | yeah | 23:31 |
| flwang | it's a good start, which can save much time of us | 23:32 |
| vkmc | indeed | 23:32 |
| vkmc | :) | 23:32 |
| vkmc | using websocket is straightforward with Javascript | 23:33 |
| vkmc | we only need to modify this a bit https://review.openstack.org/#/c/40198/4/horizon/static/horizon/js/horizon.realtime.js | 23:33 |
| vkmc | as you said | 23:33 |
| vkmc | have it listening to Zaqar | 23:33 |
| flwang | and based on the patch, seems they(horizon) don't really need to know every status change | 23:33 |
| vkmc | nope | 23:34 |
| flwang | and for a demo, it should be fine as well | 23:34 |
| vkmc | now, what I generating some noise for me | 23:34 |
| vkmc | how is the client supposed to return a request for a response generated by Horizon to Zaqar? | 23:34 |
| vkmc | or we put Zaqar in the middle and just pass the request through Zaqar | 23:35 |
| vkmc | I think I just answered myself | 23:35 |
| flwang | why there is a request? | 23:35 |
| vkmc | a user creates an instance through Horizon | 23:35 |
| vkmc | then a request for Nova is created | 23:36 |
| vkmc | its issued | 23:36 |
| flwang | no matter the request is created by horizon or not | 23:36 |
| flwang | it doesn't matter | 23:36 |
| vkmc | but but | 23:36 |
| vkmc | k | 23:36 |
| * vkmc tends to overthink stuff | 23:36 | |
| flwang | the scenario maybe like this: a user is viewing the instances page, another user create a new instance by cli | 23:37 |
| *** Qiming has joined #openstack-zaqar | 23:37 | |
| flwang | nova got the request from the nova client(cli), send notification to ceilometer, if we have a publisher to zaqar, then the notification will be forwarded to zaqar's queue, zaqar notify horizon by websocket(therve's patch), update the table | 23:38 |
| vkmc | oki | 23:40 |
| vkmc | sounds good to me | 23:40 |
| flwang | the notification path from nova(or other serivce) to zaqar depends on how we implement it | 23:40 |
| flwang | for demo, you can use a tool like therve mentioned, but in openstack world, leveraging existing service would make more sense, thoughts? | 23:42 |
| flwang | i dunno, we can discuss when we demo it, obviously, it's not a blocker for now :) | 23:43 |
| vkmc | certainly yes | 23:43 |
| vkmc | we only want to make sure we can show that Horizon won't need to be continuously polling n services to be aware of status changes | 23:43 |
| flwang | exactly | 23:43 |
| vkmc | so in the demo we should show how its implemented now | 23:43 |
| vkmc | and how it should be improved by Zaqar usage | 23:44 |
| flwang | vkmc: this demo can show the two new features of zaqar, your websocket and the notifications | 23:44 |
| flwang | it would be really cool | 23:44 |
| vkmc | yes :D | 23:45 |
| flwang | oh, btw, i will take vacation since next Monday until the summit :) | 23:45 |
| flwang | but | 23:45 |
| flwang | i will be around | 23:45 |
| flwang | and my timezone will be UTC+8 | 23:46 |
| vkmc | oh oki | 23:46 |
| flwang | i'm so excited for this demo | 23:47 |
| vkmc | going back to your homeland for the holidays? :) | 23:47 |
| vkmc | yeah me too | 23:47 |
| vkmc | I'll try to get things done with time so we can see it in action with time | 23:48 |
| flwang | yes, back to Beijing to join my young brother's wedding and see my parents | 23:48 |
| vkmc | ohh, congrats! | 23:48 |
| flwang | besides, i'm also thinking another demo about notifications or the basic messaging service of zaqar, like https://aws.amazon.com/blogs/aws/auto-scaling-with-sqs/ and http://docs.aws.amazon.com/AutoScaling/latest/DeveloperGuide/as-using-sqs-queue.html | 23:51 |
| *** Qiming has quit IRC | 23:52 | |
| vkmc | oh yes, I was going through those examples the other day | 23:53 |
| flwang | now the ^only^ doc about auto scaling of openstack is this one http://techs.enovance.com/7437/autoscaling-with-heat-ceilometer-gnocchi | 23:53 |
| vkmc | and it caught my attention the concept of "invisibility" for messages | 23:53 |
| flwang | but, it's only based on the hardware performance | 23:53 |
| vkmc | has that been discussed in Marconi/Zaqar in the past? | 23:53 |
| flwang | IMHO, it's not really reasonable | 23:53 |
| flwang | the autoscaling should be based on the application workload instead hardware workload | 23:53 |
| * vkmc checks the links | 23:54 | |
| flwang | vkmc: i don't think so? | 23:54 |
| vkmc | what exaclty "hardware workload" means? | 23:54 |
| flwang | like the cpu util is more than 80% or something like that | 23:55 |
| flwang | memory util > 80% | 23:55 |
| vkmc | so how much of the available resources are being used | 23:55 |
| vkmc | instead of how many processes are being executed | 23:56 |
| vkmc | which is application workload | 23:56 |
| vkmc | its hard to say | 23:56 |
| flwang | vkmc: but, for distributed application(like most of the web application), the message count in the queue(or similar) is the key part they should be aware, and which is the real workload for the application | 23:56 |
| vkmc | well yes, that's true | 23:57 |
| flwang | but yes, if it's not really based on message queue, then it maybe not true | 23:57 |
| flwang | i would like to show more user scenarios | 23:58 |
| vkmc | yeah, I think that's something we are missing in Zaqar right now | 23:59 |
| vkmc | showing the potential Zaqar has will help us gain adoption | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!