| *** harlowja has quit IRC | 00:09 | |
| *** kgriffs has quit IRC | 00:12 | |
| *** kgriffs has joined #openstack-zaqar | 00:17 | |
| *** harlowja has joined #openstack-zaqar | 00:25 | |
| *** mpanetta has joined #openstack-zaqar | 00:36 | |
| *** mpanetta has quit IRC | 00:41 | |
| *** wanghao has joined #openstack-zaqar | 00:43 | |
| openstackgerrit | Merged openstack/zaqar: Use _LW() to wrap the warning message https://review.openstack.org/411234 | 01:23 |
|---|---|---|
| openstackgerrit | Merged openstack/zaqar: Fix typos in HACKING.rst https://review.openstack.org/416803 | 01:23 |
| *** yanyanhu has joined #openstack-zaqar | 01:32 | |
| *** lhx__ has joined #openstack-zaqar | 01:33 | |
| *** nadya has joined #openstack-zaqar | 01:51 | |
| *** nadya has quit IRC | 01:55 | |
| *** nadya has joined #openstack-zaqar | 03:51 | |
| *** lhx__ has quit IRC | 03:52 | |
| *** nadya has quit IRC | 03:56 | |
| *** flwang1 has quit IRC | 03:59 | |
| *** lhx__ has joined #openstack-zaqar | 04:03 | |
| *** nadya has joined #openstack-zaqar | 04:52 | |
| *** nadya has quit IRC | 04:56 | |
| *** lhx__ has quit IRC | 05:13 | |
| *** nadya has joined #openstack-zaqar | 05:32 | |
| *** nadya has quit IRC | 05:34 | |
| *** lhx__ has joined #openstack-zaqar | 06:01 | |
| *** nadya has joined #openstack-zaqar | 06:07 | |
| *** nadya has quit IRC | 06:11 | |
| *** itisha has quit IRC | 06:12 | |
| *** pcaruana has joined #openstack-zaqar | 06:25 | |
| *** tesseract has joined #openstack-zaqar | 07:08 | |
| *** nadya has joined #openstack-zaqar | 07:11 | |
| *** lhx__ has quit IRC | 08:19 | |
| *** Eva-i has quit IRC | 08:22 | |
| *** Eva-i has joined #openstack-zaqar | 08:25 | |
| *** lhx__ has joined #openstack-zaqar | 08:31 | |
| *** wanghao has quit IRC | 09:28 | |
| *** nadya has quit IRC | 09:45 | |
| *** AAzza has quit IRC | 09:56 | |
| *** AAzza has joined #openstack-zaqar | 09:57 | |
| *** nadya has joined #openstack-zaqar | 10:05 | |
| *** nadya has quit IRC | 10:22 | |
| openstackgerrit | avnish proposed openstack/python-zaqarclient: Removed unnecessary utf-8 encoding https://review.openstack.org/418824 | 10:31 |
| *** nadya has joined #openstack-zaqar | 10:50 | |
| *** nadya has quit IRC | 11:03 | |
| *** yanyanhu has quit IRC | 11:24 | |
| *** nadya has joined #openstack-zaqar | 11:50 | |
| *** rbrady-afk is now known as rbrady | 12:50 | |
| *** AAzza has quit IRC | 13:06 | |
| *** AAzza has joined #openstack-zaqar | 13:07 | |
| *** lhx__ has quit IRC | 13:08 | |
| *** AAzza has quit IRC | 13:13 | |
| *** AAzza has joined #openstack-zaqar | 13:14 | |
| *** AAzza has quit IRC | 13:18 | |
| *** AAzza has joined #openstack-zaqar | 13:19 | |
| *** lhx__ has joined #openstack-zaqar | 13:45 | |
| *** david-lyle has quit IRC | 13:56 | |
| *** david-lyle has joined #openstack-zaqar | 13:59 | |
| *** itisha has joined #openstack-zaqar | 14:11 | |
| *** david-lyle has quit IRC | 14:16 | |
| *** mpanetta has joined #openstack-zaqar | 14:43 | |
| *** mpanetta has quit IRC | 14:48 | |
| *** mpanetta has joined #openstack-zaqar | 14:49 | |
| *** lhx__ has quit IRC | 15:04 | |
| *** AAzza has quit IRC | 15:16 | |
| *** AAzza has joined #openstack-zaqar | 15:16 | |
| *** adu has joined #openstack-zaqar | 15:24 | |
| *** lhx__ has joined #openstack-zaqar | 15:48 | |
| *** wxy| has joined #openstack-zaqar | 15:50 | |
| *** lhx__ has quit IRC | 16:04 | |
| dtrainor | Hi. I'm working on proxying the services that UI requires with the Apache instance that's serving the UI. I've managed to proxy all the services except Zaqar. This may be for a few different reasons, including my inexperience with how websockets operate over http. An example of the Apache configuration I'm prototyping is located at http://paste.openstack.org/show/594589/ . I've also made sure that mod_proxy, mod_proxy_http | 16:50 |
| dtrainor | , and mod_proxy_wstunnel are loaded. The error I'm getting is "ZaqarWebSocketService.js:16 WebSocket connection to 'wss://8.43.87.242/zaqar-ws' failed: Error during WebSocket handshake: Unexpected response code: 200". Did some research on this one but I can't find anything related to autobahn, the websocket implementation we're using for UI. | 16:50 |
| wxy| | dtrainor, hi, is there any error log in zaqar-websocket? | 16:53 |
| wxy| | dtrainor: And could you please open a bug for this? | 16:54 |
| wxy| | dtrainor: basically, response code 200 means that the handshake is sucesse. It's a little strange for your situation. | 17:00 |
| dtrainor | hello, sorry, got distratected for a moment. working on pulling some more logs | 17:00 |
| dtrainor | i had read that as well | 17:00 |
| *** nadya has quit IRC | 17:01 | |
| dtrainor | however I don't see any errors (or connection attempts) in the zaqar debug logs, for the wsgi nor the websocket server. i haven't opened a bug because this is a feature i'm working on. | 17:01 |
| dtrainor | there's a blueprint associated with it | 17:01 |
| wxy| | dtrainor: So I guess the problem is not in zaqar inside. it's between Apache and Zaqar. | 17:03 |
| dtrainor | i wish i could say that with confidence, yes, but my limited knowledge of how websockets work makes me hesitant to point that out exclusively | 17:04 |
| dtrainor | it's slowly coming together in my brain, and troubleshooting implementations like this help | 17:04 |
| wxy| | dtrainor: You can register a bug in Zaqar launchpad first. I'm sure we'll take a look and help you a few hours later. My TZ is UTC +8, it's 1:00 a.m. I have to go sleep now. | 17:08 |
| dtrainor | oh, i haven't yet turned on apache debug logging, taking a look at that | 17:08 |
| dtrainor | oh for sure. thanks. | 17:08 |
| wxy| | dtrainor: :), good luck. | 17:08 |
| dtrainor | thank you for the time | 17:08 |
| *** wxy| has quit IRC | 17:09 | |
| *** pcaruana has quit IRC | 17:12 | |
| *** nadya has joined #openstack-zaqar | 17:14 | |
| *** wxy has quit IRC | 17:14 | |
| *** tesseract has quit IRC | 17:20 | |
| *** nadya has quit IRC | 17:24 | |
| *** rbrady has quit IRC | 17:46 | |
| *** nadya has joined #openstack-zaqar | 17:50 | |
| *** vipuls has joined #openstack-zaqar | 17:58 | |
| *** nadya has quit IRC | 18:01 | |
| *** david-lyle has joined #openstack-zaqar | 18:16 | |
| *** nadya has joined #openstack-zaqar | 18:35 | |
| *** wxy has joined #openstack-zaqar | 19:12 | |
| *** rbrady has joined #openstack-zaqar | 19:52 | |
| *** rbrady has quit IRC | 19:52 | |
| *** rbrady has joined #openstack-zaqar | 19:52 | |
| flwang | dtrainor: still around? | 20:02 |
| dtrainor | yep sure am. | 20:02 |
| dtrainor | i may have figured it out, but I believe my ProxyPassReverse may need tweaking | 20:03 |
| flwang | dtrainor: ok, cool | 20:05 |
| flwang | feel free ask questions here, we're always around | 20:05 |
| dtrainor | appreciate it, thanks. what i'm seeing is the client (ui) make that websocket connection, but not actually receiving any data | 20:06 |
| flwang | dtrainor: btw, do you have any doc/blog to track the zaqar usage based on your experience? | 20:06 |
| dtrainor | just some scratch notes, nothing composed. this would be a good opportunity to create some though. | 20:06 |
| *** nadya has quit IRC | 20:07 | |
| flwang | dtrainor: yep, you can take this a kind request from zaqar team, tripleO is a good consumer of zaqar, and it would be nice if you guys can write some docs based on your experience, which will benefit the others a lot, i'm sure :) | 20:07 |
| dtrainor | sounds liek a good opportunity all around | 20:08 |
| flwang | dtrainor: thank you in advance ;) | 20:12 |
| dtrainor | my pleasure | 20:12 |
| dtrainor | these last few days are the first i've ever taken to look in to zaqar, still coming up to speed | 20:12 |
| *** flwang1 has joined #openstack-zaqar | 20:14 | |
| dtrainor | (in case you didn't see it) these last few days are the first i've ever taken to look in to zaqar, still coming up to speed | 20:15 |
| flwang | all good, that could help us to fix some UX issue :) | 20:20 |
| flwang | i mean make the user easy to use zaqar | 20:20 |
| dtrainor | ok, so here's a question. i see in apache debug logs that a request is made via proxy to the zaqar websocket instance but my client (the ui) never receives any data | 20:25 |
| dtrainor | that is, things that rely on zaqar ws to be updated in the ui never change | 20:25 |
| dtrainor | that's why i suspected problems with ProxyPassReverse | 20:25 |
| dtrainor | though, since it's a websocket, i'm not certain i even need a ProxyPassReverse.... zaqar doesn't provide any kind of redirects or anything, does it? | 20:27 |
| flwang | i don't think zaqar provides any redirect | 20:38 |
| flwang | unless there is a builtin support from falcon | 20:38 |
| flwang | may i know more background of your feature/test? | 20:39 |
| flwang | currently, devstack is using uwsgi as the wsgi server, seems you're using mod_wsgi, right? | 20:39 |
| flwang | and then you have haproxy before the mod_wsgi, is that your architecture? | 20:40 |
| dtrainor | if i understand this correctly, the wsgi server runs but also invokes an additional transport for websockets | 20:42 |
| flwang | can you talk to zaqar without the haproxy? | 20:43 |
| dtrainor | my platform is the osp10 product but for this, i think it's close enough to devstack and upstream that the problem should exist in these different product variations | 20:43 |
| dtrainor | that's a good question, i juts assumed "yes", let me see about testing that | 20:43 |
| dtrainor | i see in the zaqar debug/verbose logs that requests are coming in | 20:44 |
| dtrainor | 2017-01-11 15:41:44.166 32498 INFO zaqar.transport.websocket.protocol [-] Response: API v2 txt, 201. Request: action "subscription_create", body {"queue_name": "tripleo", "ttl": 3600}. | 20:44 |
| dtrainor | well, that a subscription to the queue happens (if i'm reading this properly) presumably by the client (the ui) | 20:44 |
| dtrainor | ok, i can curl zaqar (https://wiki.openstack.org/wiki/Zaqar/Give_Zaqar_a_try) :8888 and get an unauthorized, and curl'ing :9000 (the websocket transport listening port) and get an "curl: (52) Empty reply from server" | 20:50 |
| dtrainor | the only part of zaqar that ui uses is the websocket transport | 20:50 |
| dtrainor | [Wed Jan 11 15:56:58.530276 2017] [proxy_http:debug] [pid 8108] mod_proxy_http.c(1925): [client 192.0.2.1:60742] AH01113: HTTP: declining URL ws://192.0.2.1:9000/ | 20:57 |
| dtrainor | is there a special part of a url i'm missing when sending connections to that websocket transport? do I need to include /websocket or something in the uri? i'm seeing a few posts that suggest this sometimes may apply | 21:01 |
| *** AAzza has quit IRC | 22:04 | |
| *** AAzza has joined #openstack-zaqar | 22:05 | |
| *** AAzza has quit IRC | 22:09 | |
| *** AAzza has joined #openstack-zaqar | 22:11 | |
| *** AAzza has quit IRC | 22:20 | |
| *** AAzza has joined #openstack-zaqar | 22:30 | |
| *** AAzza has quit IRC | 22:41 | |
| *** AAzza has joined #openstack-zaqar | 22:43 | |
| flwang | dtrainor: sorry for the late response. was in a long meeting | 22:54 |
| dtrainor | no worries - i was mostly rambling | 22:54 |
| flwang | when you run curl, did you set the correct token? | 22:54 |
| dtrainor | i didn't use curl to generate a token, just tried to curl the url without authentication. i was looking at some other angles, and found that apache may upgrade the connection for websockets when using mod_proxy_wstunnel, and that connection gets upgraded for the entireity of the request, which may cause problems such as thi | 22:55 |
| dtrainor | *this | 22:55 |
| dtrainor | https://lists.apache.org/thread.html/c0d000d85f5afe1a7d7e0e6fef5c8d6b891dc8483814c567c593bf35@%3Cusers.httpd.apache.org%3E | 22:55 |
| dtrainor | I think this is more of an Apache problem | 22:56 |
| dtrainor | I have another environment that's identical, except it's not using the proxying technique that i'm working on. So I can reproduce happy path or sad path tests to compare against. | 22:56 |
| dtrainor | i need to leave to go meet a coworker for dinner | 22:58 |
| dtrainor | i really appreciate your help and ideas, thank you. i'll be back at it tonight and tomorrow. | 22:58 |
| flwang | cool, ttyl | 23:05 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!