| *** wxy has joined #openstack-zaqar | 01:04 | |
| *** david_cu has joined #openstack-zaqar | 04:55 | |
| *** njohnston has quit IRC | 05:02 | |
| *** njohnston has joined #openstack-zaqar | 05:04 | |
| *** david_cu has quit IRC | 05:18 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/zaqar: Imported Translations from Zanata https://review.openstack.org/307595 | 06:23 |
|---|---|---|
| *** rcernin has joined #openstack-zaqar | 07:13 | |
| *** pcaruana has joined #openstack-zaqar | 10:32 | |
| *** AAzza has quit IRC | 11:08 | |
| *** AAzza has joined #openstack-zaqar | 11:09 | |
| *** pcaruana has quit IRC | 13:28 | |
| *** amitgandhinz has joined #openstack-zaqar | 14:03 | |
| adu | flwang: pong | 14:58 |
| adu | I tried to run the wsgi and got: 2016-04-21 11:34:34.243 3548 CRITICAL zaqar [-] InvalidDriver: No 'zaqar.control.storage' driver found, looking for 'mongodb' | 14:59 |
| *** pcaruana has joined #openstack-zaqar | 15:21 | |
| *** Eva-i has quit IRC | 15:22 | |
| *** Eva-i has joined #openstack-zaqar | 15:28 | |
| *** mpanetta has joined #openstack-zaqar | 16:06 | |
| *** GB21 has joined #openstack-zaqar | 16:19 | |
| *** david_cu has joined #openstack-zaqar | 16:37 | |
| *** david_cu_ has joined #openstack-zaqar | 16:38 | |
| *** david_cu has quit IRC | 16:41 | |
| *** shakamunyi has joined #openstack-zaqar | 16:59 | |
| *** david_cu_ has quit IRC | 17:02 | |
| *** tesseract has joined #openstack-zaqar | 17:06 | |
| *** tesseract is now known as Guest8613 | 17:06 | |
| Eva-i | ryansb: vkmc: hello, people. How the summit is going? | 17:07 |
| Eva-i | ryansb: vkmc: have you decided who's going to provide laptop for me and flwang to attend to the design sessions remotely? | 17:10 |
| *** GB21 has quit IRC | 17:14 | |
| *** GB21 has joined #openstack-zaqar | 17:51 | |
| *** GB21 has quit IRC | 18:24 | |
| *** david-lyle has joined #openstack-zaqar | 18:25 | |
| *** GB21 has joined #openstack-zaqar | 18:35 | |
| *** rcernin has quit IRC | 18:47 | |
| *** david-lyle has quit IRC | 18:49 | |
| *** GB21 has quit IRC | 18:52 | |
| *** GB21 has joined #openstack-zaqar | 18:55 | |
| *** david-lyle has joined #openstack-zaqar | 19:00 | |
| *** GB21 has quit IRC | 19:06 | |
| *** flwang1 has joined #openstack-zaqar | 19:09 | |
| flwang1 | adu: ping | 19:09 |
| adu | flwang1: are you a bot? | 19:09 |
| flwang1 | adu: no :) | 19:09 |
| adu | yey | 19:09 |
| flwang1 | yesterday, you asked how to deploy zaqar in productioin, right? | 19:10 |
| adu | yes | 19:10 |
| flwang1 | adu: may i know more details? I'm the PTL of zaqar, hopefully i can help | 19:10 |
| adu | I’ve followed the instruction to setup a dev server, with backend=noauth | 19:10 |
| *** david-lyle has quit IRC | 19:10 | |
| adu | and I’m using python-zaqarclient to construct a client object with the following conf | 19:11 |
| adu | http://pastie.org/10814153 | 19:11 |
| adu | I | 19:12 |
| adu | I’ve figured out how to create and delete messages | 19:12 |
| flwang1 | adu: so what's the error you're seeing? | 19:12 |
| flwang1 | adu: ok, cool | 19:12 |
| adu | and I’ve figured out how to claim messages with curl, but I can’t figure out how to claim messages with python-zaqarclient | 19:13 |
| flwang1 | adu: i'm curious why you want to use noauth mode? | 19:13 |
| adu | when i do client.queue(‘Q1’).claim(ttl=1209500, grace=60, limit=1) I get: | 19:13 |
| adu | MalformedRequest: Error response from Zaqar. Code: 400. Title: Invalid API request. Description: The TTL for a claim may not exceed 1209600 seconds, and must be at least 60 seconds long.. | 19:13 |
| adu | I don’t want to use noauth mode, I want to use keystone, but I only have a finite amount of time in my day... | 19:14 |
| flwang1 | :) can't see the relationship between the auth mode and your time of day | 19:15 |
| flwang1 | did you mean the auth mode will take more time to set up? | 19:15 |
| adu | possibly, I don’t know | 19:15 |
| flwang1 | adu: ok, if it's a bug, i can help figure out and fix. but before that, would you mind sharing more info about your side? | 19:17 |
| adu | my side? | 19:17 |
| adu | hmm, as in mongodb? | 19:17 |
| flwang1 | like what's your usercase? why do you want to deploy zaqar in your prod? | 19:17 |
| adu | ah | 19:17 |
| adu | order processing | 19:18 |
| flwang1 | so you guys have already deployed openstack, right? which version? | 19:18 |
| adu | we currently have about 6 distinct queue systems, most of which are for one-off jobs, and zaqar is a nice fit for that, I think | 19:19 |
| adu | no, we have a contract with Rackspace | 19:19 |
| flwang1 | it's a PoC or you guys have decided to use zaqar? | 19:19 |
| adu | and that’s pretty similar to OpenStack | 19:19 |
| flwang1 | adu: so Rackspace will help you guys deploy openstack in your prod? | 19:20 |
| adu | I wanted to compare Rackspace Cloud Queues with an in-house Zaqar deployment | 19:20 |
| flwang1 | ah, i see. | 19:21 |
| mpanetta | adu: Zaqar would be better. Cloud Queues is in sustain mode. | 19:21 |
| adu | flwang1: no, iirc, rackspace is to openstack as netscape is to mozilla | 19:21 |
| mpanetta | I am the maintainer of CQ which is why I know this heh | 19:21 |
| flwang1 | mpanetta: haha, thanks for showing up :) | 19:22 |
| mpanetta | I'm always here :P Just in hiding hehe | 19:22 |
| flwang1 | adu: Rackspace is using the early version of zaqar as their cloud queues | 19:22 |
| adu | mpanetta: So can I use an in-house zaqar that points to rackspace keystone? or would I have to run our own keystone server? | 19:22 |
| mpanetta | yes quite an early version | 19:22 |
| mpanetta | I'm thinking you would have to set up your own keystone | 19:23 |
| flwang1 | adu: if you're concerning about the performance, i think an in-house deployment is better for sure | 19:23 |
| mpanetta | Since you don't have admin rights to the rackspace one. | 19:23 |
| mpanetta | Yes it most definitely is better | 19:23 |
| flwang1 | adu: and IIRC, Rackspace is not using Keystone as auth | 19:24 |
| flwang1 | mpanetta: pls fix me ^ | 19:24 |
| mpanetta | They aren't | 19:24 |
| mpanetta | It is a contract compliant thing | 19:24 |
| mpanetta | but not keystone proper | 19:24 |
| adu | I thought it was keystone with a million custom extensions | 19:25 |
| mpanetta | nah | 19:25 |
| mpanetta | its a java app of some sort iirc | 19:25 |
| adu | well, that’s not very pythonic :/ | 19:25 |
| mpanetta | hah yeah | 19:25 |
| flwang1 | adu: so based on what we discussed above, it's totally depending on your user case | 19:26 |
| mpanetta | adu: If it is private cloud I would set up my own probably. | 19:26 |
| adu | so I already see how zaqar can help with most of our order processing needs, but there is one area that we need help with that isn’t so obvious | 19:27 |
| flwang1 | if you really want a in-house queue to get better performance, then zaqar is a good choice, but if you want it to play with Rackspace cloud, you may need some extra effort | 19:27 |
| flwang1 | adu: the area is ? | 19:27 |
| adu | account polling | 19:27 |
| adu | we poll customer accounts like every 2 minutes or so | 19:28 |
| flwang1 | adu: why do you guys have to use poll? | 19:28 |
| flwang1 | adu: we have notification, maybe it could help if i can know more details about t the case | 19:29 |
| adu | flwang1: that’s a good question, I’ve been asking Ticketmaster for a while now, but so far no response | 19:29 |
| flwang1 | in other words, why do you guys want to poll the account? :) | 19:29 |
| flwang1 | adu: ok, again, is it still in PoC? | 19:29 |
| flwang1 | i mean the thing you're doing :) | 19:30 |
| adu | flwang1: proof-of-concept | 19:30 |
| adu | ? | 19:30 |
| flwang1 | yes :) sorry | 19:30 |
| adu | Yes, I’m currently in proof-of-concept mode, but I was worried that there was no production user guide, only development user guide | 19:31 |
| adu | I’m trying to see if I can replace our 6 wierd queues with zaqars | 19:32 |
| flwang1 | adu: pls blame the bad doc policy of openstack | 19:32 |
| flwang1 | adu: we do have user guide | 19:32 |
| adu | ok | 19:32 |
| flwang1 | and based on the old policy of openstack, the non-core project can't put user guide on docs.openstack.org | 19:33 |
| flwang1 | and that's what i'm working on | 19:33 |
| flwang1 | adu: i will push the user guide to publish asap to address your concern | 19:35 |
| flwang1 | adu: FWIW, see https://github.com/openstack/zaqar/tree/master/doc/user-guide | 19:35 |
| adu | I think I found a rendering of that somewhere | 19:37 |
| flwang1 | adu: it's possible, but we will improve it as well | 19:39 |
| flwang1 | adu: btw, can you show me the error about the claim issue? i mean the server side log | 19:39 |
| *** david-lyle has joined #openstack-zaqar | 19:41 | |
| *** pcaruana has quit IRC | 19:45 | |
| adu | 2016-04-26 15:15:00.371 13888 DEBUG zaqar.transport.wsgi.v1_1.claims [(None,) - - - - -] [project_id:dd75a015-b93e-4a20-a715-7185a215b64f] The TTL for a claim may not exceed 1209600 seconds, and must be at least 60 seconds long. on_post /Users/ajr/1t/src/zaqar/zaqar/transport/wsgi/v1_1/claims.py:97 | 19:45 |
| adu | I put lots of LOG.debug things everywhere, and verified that _params is {'limit': '1'} and metadata is {'grace': 60, 'ttl': 1209500} | 19:45 |
| flwang1 | adu: cool, thanks, will review the source code to see if there is a bug | 19:47 |
| adu | so the claim creation in the client is passing the right data to the server | 19:47 |
| *** flwang1 has quit IRC | 19:52 | |
| *** Guest8613 has quit IRC | 19:54 | |
| *** david-lyle has quit IRC | 19:57 | |
| *** vipuls has quit IRC | 20:08 | |
| *** david_cu has joined #openstack-zaqar | 20:35 | |
| *** david_cu has quit IRC | 20:43 | |
| *** adu has quit IRC | 20:58 | |
| *** flwang1 has joined #openstack-zaqar | 21:13 | |
| *** amitgandhinz has quit IRC | 21:18 | |
| *** amitgandhinz has joined #openstack-zaqar | 21:19 | |
| *** amitgandhinz has quit IRC | 21:19 | |
| *** khushbu has joined #openstack-zaqar | 22:01 | |
| *** david_cu has joined #openstack-zaqar | 22:10 | |
| *** adu has joined #openstack-zaqar | 22:17 | |
| adu | flwang1: back | 22:18 |
| adu | um, I mean ping | 22:18 |
| flwang1 | adu: cool | 22:18 |
| adu | I just got home from work | 22:19 |
| flwang1 | adu: is that ok for you? | 22:19 |
| flwang1 | overtime ? ;) | 22:19 |
| *** khushbu has quit IRC | 22:19 | |
| adu | oh yes, since I'm having fun too :) | 22:19 |
| *** khushbu_ has joined #openstack-zaqar | 22:20 | |
| flwang1 | awesome | 22:21 |
| adu | ok, I don't have the error message anymore | 22:21 |
| flwang1 | adu: you mean the claim issue can't be reproduced anymore? | 22:21 |
| adu | oh, I can just ssh into work, n/m | 22:22 |
| flwang1 | ok, so you can still reproduce the claim issue, right? | 22:27 |
| flwang1 | just double check | 22:27 |
| adu | yes | 22:27 |
| adu | I just need to ssh to reproduce it :P | 22:27 |
| adu | Sorry for the confusion | 22:28 |
| flwang1 | ok, got it | 22:28 |
| adu | I wonder if that issue would go away with keystone | 22:28 |
| flwang1 | adu: would you mind opening a bug https://bugs.launchpad.net/zaqar/+filebug ? | 22:28 |
| adu | like if the validation is failing because x-auth-token is missing, but it raises a TTL issue instead? | 22:28 |
| flwang1 | to track this issue and it would be nice if you can add some reproduce steps | 22:29 |
| flwang1 | then today i will work on that | 22:29 |
| flwang1 | and as for your user case, i think we're good. either TTL or delete manually could work | 22:30 |
| flwang1 | adu: i assume you know how to set TTL of message, and as for message delete, here is the api https://wiki.openstack.org/wiki/Zaqar/specs/api/v1.1#Delete_Multiple_Messages | 22:31 |
| *** david_cu has quit IRC | 22:32 | |
| *** david_cu has joined #openstack-zaqar | 22:36 | |
| adu | ok | 22:39 |
| flwang1 | so do you have any other concern? | 22:40 |
| flwang1 | about using zaqar to meet your user case | 22:40 |
| adu | concerns? | 22:44 |
| adu | I'm not concerned about this particular issue | 22:44 |
| adu | I'm pretty sure it's me being a dumb ass | 22:44 |
| adu | My only concerns are that zaqar can take care of one-off and periodic jobs | 22:45 |
| flwang1 | yep, that is what i'm asking :) | 22:45 |
| flwang1 | sorry, i'm not a native english speaker | 22:46 |
| adu | flwang1: what is your native language? | 22:46 |
| flwang1 | mandarin | 22:47 |
| adu | oOo | 22:47 |
| flwang1 | I"m a Chinese | 22:47 |
| *** mpanetta has quit IRC | 22:51 | |
| adu | So another thing is, I was running both zaqar-server and python-zaqarclient on MacOSX | 22:54 |
| adu | but when I try and reproduce it in linux, I get: "ImportError: cannot import name IncompleteRead" when trying to pip install python-zaqarclient | 22:55 |
| adu | I think I remember pip install -U pip fixing this | 22:55 |
| flwang1 | adu: did you every try use devstack to get a fully work env? | 22:56 |
| adu | flwang1: yes, my computer crashed because of not enough RAM | 22:57 |
| flwang1 | wow | 22:57 |
| adu | it has 16 GB | 22:57 |
| flwang1 | hmm.... i would recommend install devstack in a VM if so | 22:58 |
| flwang1 | with ubuntu14.04 | 22:58 |
| adu | it was in a vm | 23:00 |
| adu | and I think I allocated 8GB of ram to the vm | 23:00 |
| adu | it still crashed | 23:00 |
| adu | but then again, I think I had 25 services configured | 23:00 |
| flwang1 | adu: hmm... interesting, it works fine on my laptop and VM | 23:00 |
| flwang1 | ah, why not try disable most of the other service and only keep keystone, zaqar and the others you want | 23:01 |
| flwang1 | anyway, i don't think 'any bug' is a problem to block you to use zaqar in your production | 23:01 |
| adu | I don't think I tried just keystone and zaqar | 23:02 |
| flwang1 | so i would suggest let's focus on the user cases | 23:02 |
| adu | user cases? | 23:02 |
| flwang1 | yep, i mean what you're trying: if zaqar can meet all your requirements | 23:02 |
| adu | yes | 23:03 |
| *** khushbu_ has quit IRC | 23:03 | |
| flwang1 | so anything i can help to make things easier? | 23:05 |
| flwang1 | accelerate your test, etc, etc :) | 23:05 |
| adu | I guess I'll use devstack | 23:09 |
| adu | I just don't understand devstack too well, which is why I wanted to run it in isolation | 23:10 |
| adu | I think I tried adding zaqar to the default devstack, and I still got 25 services running | 23:10 |
| flwang1 | do you want to try it today? | 23:12 |
| flwang1 | if so,i can share my localrc file for you | 23:12 |
| flwang1 | so that you can give it a try right now | 23:12 |
| adu | oh, that would help | 23:13 |
| flwang1 | wait a sec | 23:15 |
| flwang1 | adu: http://paste.openstack.org/show/495474/ | 23:16 |
| flwang1 | if you're using a VM, replace the SERVICE_HOST and HOST_IP with your VM's IP address | 23:16 |
| flwang1 | so that you can talk with those openstack service from your host | 23:16 |
| adu | flwang1: I think it was glance and cinder that was causing the crash, do you have one with those services disabled? | 23:17 |
| flwang1 | adu: you can comment them out, but i would suggest try again and show me the errors | 23:18 |
| flwang1 | i guess it's cinder | 23:18 |
| flwang1 | so you can try comment out cinder first | 23:18 |
| adu | ok, will do, I will update in several hours | 23:23 |
| flwang1 | adu: awesome, btw, what's your timezone? | 23:23 |
| adu | -5 | 23:25 |
| adu | I think | 23:25 |
| adu | EST | 23:25 |
| adu | "EST5EDT" in tzdata | 23:25 |
| flwang1 | i always use this http://www.worldtimebuddy.com/ | 23:29 |
| adu | I use timeanddate.com | 23:38 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!