Eva-i | flwang1: now I checked Zaqar and can't find problem... I don't know on which side the problem is... | 00:01 |
---|---|---|
*** eliqiao has quit IRC | 00:22 | |
*** eliqiao has joined #openstack-zaqar | 00:23 | |
openstackgerrit | Thai Tran proposed openstack/zaqar-ui: Zaqar-ui base project with empty queues panel https://review.openstack.org/255213 | 00:34 |
flwang1 | Eva-i: i'm thinking if there is a problem with the zaqar client code | 00:58 |
openstackgerrit | Thai Tran proposed openstack/zaqar-ui: Zaqar-ui base project with empty queues panel https://review.openstack.org/255213 | 01:00 |
*** kgriffs is now known as kgriffs|afk | 01:17 | |
*** eliqiao has quit IRC | 01:24 | |
*** eliqiao has joined #openstack-zaqar | 01:42 | |
flwang1 | eliqiao: hah, welcome to visit zaqar channel ;) | 01:54 |
eliqiao | hi flwang1, thx :) actually I lived here for sometime already(dive into water) | 01:55 |
Eva-i | eliqiao: hello | 02:01 |
eliqiao | hi Eva-i, yes? | 02:01 |
Eva-i | eliqiao: just hello =) | 02:01 |
eliqiao | Eva-i: :) thx | 02:02 |
flwang1 | Eva-i: not sure if you notice this line http://logs.openstack.org/82/261282/5/check/gate-zaqarclient-dsvm-functional/17d875e/console.html#_2016-01-12_03_58_01_363 | 02:06 |
Eva-i | flwang1: this happens because of another bug | 02:07 |
Eva-i | flwang1: see http://logs.openstack.org/96/238396/12/check/gate-dsvm-redis-zaqar/9616539/logs/screen-zaqar-wsgi.txt.gz#_2016-01-12_09_19_58_799 | 02:07 |
flwang1 | Eva-i: i'm thinking how the resp_dict could be empty | 02:12 |
flwang1 | if zaqar can't find the subscription_id, it should reutrn 404 | 02:12 |
flwang1 | http://logs.openstack.org/96/238396/12/check/gate-dsvm-redis-zaqar/9616539/logs/screen-zaqar-wsgi.txt.gz#_2016-01-12_09_19_58_795 | 02:14 |
Eva-i | flwang1: you think it's because no pool is found? | 02:15 |
Eva-i | flwang1: BTW I checked requests which python-zaqarclient sends to zaqar during 'subscription_get' operation using wireshark. Everything is alright. The server just acknowledges this request and sends reset connection packet: http://imgur.com/mbT0T49 | 02:16 |
*** Qiming has joined #openstack-zaqar | 02:16 | |
Eva-i | During this python-zaqarclient shows: requests.exceptions.ConnectionError: ('Connection aborted.', error(104, 'Connection reset by peer')) | 02:17 |
flwang1 | it's really weird | 02:17 |
Eva-i | flwang1: vmkc made an assumption, that this is a RACE CONDITION happening | 02:17 |
Eva-i | flwang1: I think the race condition happens on zaqar-server side | 02:18 |
flwang1 | Eva-i: but you said zaqar server even didn't get the request of subscription get, right? | 02:18 |
flwang1 | then how the race condition happened? | 02:18 |
Eva-i | flwang1: hm wait. maybe race condition in uwsgi | 02:19 |
Eva-i | flwang1: the problem completely eliminates if I put time.sleep(0.1) before this line: https://github.com/openstack/python-zaqarclient/blob/master/zaqarclient/queues/v2/subscription.py#L60 | 02:20 |
Eva-i | flwang1: I think it's a sign of race condition | 02:20 |
flwang1 | haha, that's what i saw at somwhere, let me show you | 02:21 |
flwang1 | https://github.com/Huangtuzhi/top-topic-Zhihu/issues/1 | 02:22 |
flwang1 | Eva-i: when you say "the problem completely eliminates if I put time.sleep(0.1) " how could you know it's really ok, just ok for this time you just tried? | 02:24 |
Eva-i | flwang1: I tried more than once, at least 10 times | 02:24 |
flwang1 | so you mean without the sleep, you can easily reproduce it on your local, right? | 02:25 |
Eva-i | flwang1: that's right | 02:25 |
flwang1 | after added that line, you can't reproduce it anymore, right? | 02:25 |
Eva-i | flwang1: right | 02:25 |
flwang1 | hmm... interesting.... | 02:25 |
flwang1 | but, i don't think it's a race condition issue of zaqar server | 02:26 |
Eva-i | flwang1: you are not able to reproduce connectivity errors on your local? | 02:28 |
flwang1 | i never use uwsgi for testing :( | 02:29 |
flwang1 | Eva-i: i'm thinking if it can be fixed by change the uwsgi config | 02:30 |
Eva-i | flwang1: me too, but for this case it was easy for me to start devstack, which configures uwsgi and starts zaqar | 02:31 |
flwang1 | Eva-i: yep, i see | 02:32 |
Eva-i | flwang1: what is relevant on this page: https://github.com/Huangtuzhi/top-topic-Zhihu/issues/1 ? | 02:34 |
Eva-i | The author uses sleep to prevent error? | 02:35 |
flwang1 | Eva-i: yes | 02:35 |
flwang1 | but obviously, we can't do that :( in our code | 02:35 |
Eva-i | flwang1: of course | 02:36 |
flwang1 | but good to know the hacking works | 02:37 |
*** mdnadeem has joined #openstack-zaqar | 02:51 | |
Eva-i | flwang1: I tried so many things today/tonight to make tests work. I don't think I can make them work in a reasonable amount of time. Maybe I should go back to my other tasks. To sleeping, for example. | 03:24 |
Eva-i | flwang: good bye | 03:24 |
flwang1 | hah, good night | 03:25 |
flwang1 | i will continually work on that | 03:25 |
openstackgerrit | Fei Long Wang proposed openstack/python-zaqarclient: Add support for subscription show v2 https://review.openstack.org/261282 | 03:26 |
*** flwang has quit IRC | 04:09 | |
*** GB21 has joined #openstack-zaqar | 05:01 | |
*** boris-42 has quit IRC | 05:03 | |
*** GB21 has quit IRC | 05:08 | |
*** Qiming_ has joined #openstack-zaqar | 05:30 | |
*** Qiming has quit IRC | 05:33 | |
*** GB21 has joined #openstack-zaqar | 05:37 | |
*** GB21 has quit IRC | 06:05 | |
*** GB21 has joined #openstack-zaqar | 06:24 | |
*** mdnadeem has quit IRC | 07:57 | |
*** mdnadeem has joined #openstack-zaqar | 07:57 | |
*** GB21 has quit IRC | 08:30 | |
openstackgerrit | Thomas Herve proposed openstack/zaqar: DNM: don't use uwsgi on the default port https://review.openstack.org/266577 | 08:39 |
*** venkat has joined #openstack-zaqar | 09:00 | |
*** GB21 has joined #openstack-zaqar | 09:25 | |
openstackgerrit | MD NADEEM proposed openstack/python-zaqarclient: Resolve asymmetric behaviour of zaqar client https://review.openstack.org/266808 | 09:33 |
openstackgerrit | Thomas Herve proposed openstack/zaqar: DNM: don't use uwsgi on the default port https://review.openstack.org/266577 | 09:38 |
openstackgerrit | MD NADEEM proposed openstack/python-zaqarclient: Resolve asymmetric behaviour of zaqar client https://review.openstack.org/266822 | 10:07 |
*** Qiming_ has quit IRC | 10:19 | |
openstackgerrit | MD NADEEM proposed openstack/zaqar: PUT improvement for pool create https://review.openstack.org/238006 | 10:26 |
openstackgerrit | MD NADEEM proposed openstack/zaqar: PUT improvement for flavor create https://review.openstack.org/238396 | 10:28 |
openstackgerrit | MD NADEEM proposed openstack/zaqar: PUT improvement for flavor create https://review.openstack.org/238396 | 10:33 |
openstackgerrit | Thomas Herve proposed openstack/zaqar: Fix redis subscription loading https://review.openstack.org/266831 | 10:40 |
therve | flwang1, So I've found one bug at least ^^^ | 10:40 |
therve | I still don't know why uswgi swallows some errors in the gate | 10:43 |
*** venkat has quit IRC | 11:11 | |
*** venkat has joined #openstack-zaqar | 11:11 | |
*** venkat has quit IRC | 11:13 | |
*** venkat has joined #openstack-zaqar | 11:13 | |
*** Qiming has joined #openstack-zaqar | 11:16 | |
*** dmowrer has quit IRC | 11:35 | |
*** dmowrer has joined #openstack-zaqar | 11:35 | |
*** flwang has joined #openstack-zaqar | 11:38 | |
*** flwang1 has quit IRC | 11:48 | |
*** flwang1 has joined #openstack-zaqar | 11:49 | |
*** mdnadeem has quit IRC | 12:18 | |
*** GB21 has quit IRC | 12:27 | |
openstackgerrit | Thomas Herve proposed openstack/zaqar: Fix redis subscription loading https://review.openstack.org/266831 | 13:07 |
*** rcernin has joined #openstack-zaqar | 13:09 | |
*** pradk has joined #openstack-zaqar | 13:19 | |
*** pradk has quit IRC | 13:23 | |
*** amitgandhinz has joined #openstack-zaqar | 13:44 | |
*** sriram has joined #openstack-zaqar | 13:59 | |
*** GB21 has joined #openstack-zaqar | 14:12 | |
*** GB21 has quit IRC | 14:12 | |
*** venkat has quit IRC | 14:16 | |
*** njohnston has joined #openstack-zaqar | 14:19 | |
*** malini has joined #openstack-zaqar | 14:29 | |
*** mdnadeem has joined #openstack-zaqar | 14:31 | |
*** malini1 has joined #openstack-zaqar | 14:32 | |
*** malini has quit IRC | 14:35 | |
*** pradk has joined #openstack-zaqar | 14:49 | |
*** Qiming has quit IRC | 14:51 | |
*** dynarro has joined #openstack-zaqar | 14:52 | |
*** mdnadeem has quit IRC | 15:06 | |
*** Qiming has joined #openstack-zaqar | 15:06 | |
*** kgriffs|afk is now known as kgriffs | 15:18 | |
*** Qiming has quit IRC | 15:20 | |
*** malini1 has quit IRC | 15:20 | |
*** boris-42 has joined #openstack-zaqar | 15:20 | |
therve | Good lord, I think I've found the uwsgi issue | 15:34 |
therve | Urrrggggggg | 15:34 |
ryansb | and the winner is? | 15:34 |
*** amitgandhinz has quit IRC | 15:35 | |
*** amitgandhinz has joined #openstack-zaqar | 15:36 | |
openstackgerrit | Thomas Herve proposed openstack/zaqar: Fix redis subscription loading https://review.openstack.org/266831 | 15:36 |
therve | ryansb, HTTP pipelining! | 15:36 |
therve | uwsgi simply closes the connection when it gets that | 15:38 |
therve | WIthout logging presumably because security | 15:38 |
therve | It's always a good reason to do something horrible to your users | 15:38 |
*** amitgandhinz has quit IRC | 16:20 | |
*** sriram has quit IRC | 16:49 | |
*** bradjones has quit IRC | 16:49 | |
*** bradjones has joined #openstack-zaqar | 16:50 | |
*** bradjones has joined #openstack-zaqar | 16:50 | |
*** rcernin has quit IRC | 16:51 | |
*** dynarro has quit IRC | 17:14 | |
*** fesp has joined #openstack-zaqar | 17:33 | |
*** fesp has quit IRC | 17:34 | |
*** amitgandhinz has joined #openstack-zaqar | 17:57 | |
*** boris-42 has quit IRC | 18:08 | |
*** wanghao has quit IRC | 18:08 | |
*** ameade has quit IRC | 18:08 | |
*** zhiyan has quit IRC | 18:08 | |
*** lane_kong has quit IRC | 18:08 | |
*** ameade has joined #openstack-zaqar | 18:21 | |
*** wanghao has joined #openstack-zaqar | 18:23 | |
*** flwang has quit IRC | 18:23 | |
*** zhiyan has joined #openstack-zaqar | 18:24 | |
*** lane_kong has joined #openstack-zaqar | 18:26 | |
*** rcernin has joined #openstack-zaqar | 18:30 | |
*** ig0r_ has joined #openstack-zaqar | 18:43 | |
*** kgriffs is now known as kgriffs|afk | 19:20 | |
*** boris-42 has joined #openstack-zaqar | 19:40 | |
*** stanchan has joined #openstack-zaqar | 20:05 | |
*** eliqiao has quit IRC | 20:17 | |
*** eliqiao has joined #openstack-zaqar | 20:20 | |
*** rcernin has quit IRC | 20:20 | |
*** kgriffs|afk is now known as kgriffs | 20:39 | |
*** rcernin has joined #openstack-zaqar | 20:49 | |
*** rcernin has quit IRC | 21:04 | |
flwang1 | therve: ping | 21:11 |
*** ig0r_ has quit IRC | 21:14 | |
*** flwang has joined #openstack-zaqar | 21:20 | |
*** ig0r_ has joined #openstack-zaqar | 21:22 | |
*** flwang has quit IRC | 21:25 | |
*** flwang has joined #openstack-zaqar | 21:29 | |
*** stanchan has quit IRC | 21:29 | |
*** mdnadeem has joined #openstack-zaqar | 21:30 | |
*** stanchan has joined #openstack-zaqar | 21:30 | |
*** ig0r_ has quit IRC | 21:31 | |
openstackgerrit | Fei Long Wang proposed openstack/python-zaqarclient: Add support for subscription show v2 https://review.openstack.org/261282 | 21:37 |
*** mdnadeem has quit IRC | 21:39 | |
*** ig0r_ has joined #openstack-zaqar | 21:44 | |
*** ig0r_ has quit IRC | 21:45 | |
openstackgerrit | Fei Long Wang proposed openstack/python-zaqarclient: Support claims for v2 https://review.openstack.org/265663 | 21:47 |
*** rcernin has joined #openstack-zaqar | 21:54 | |
*** njohnston has quit IRC | 22:01 | |
*** rcernin has quit IRC | 22:10 | |
*** malini has joined #openstack-zaqar | 22:13 | |
*** malini1 has joined #openstack-zaqar | 22:15 | |
*** malini has quit IRC | 22:17 | |
*** boris-42 has quit IRC | 22:23 | |
therve | flwang1, Yeah | 22:51 |
therve | flwang, flwang1 about to go though | 23:01 |
*** amitgandhinz has quit IRC | 23:11 | |
*** malini1 has quit IRC | 23:22 | |
*** malini has joined #openstack-zaqar | 23:23 | |
*** Qiming has joined #openstack-zaqar | 23:31 | |
*** malini1 has joined #openstack-zaqar | 23:32 | |
*** malini has quit IRC | 23:33 | |
*** Qiming has quit IRC | 23:43 | |
*** pradk has quit IRC | 23:53 | |
*** malini1 has quit IRC | 23:58 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!