| *** sbfox has quit IRC | 00:01 | |
| *** cody-somerville has quit IRC | 00:01 | |
| *** ramishra has joined #openstack-horizon | 00:10 | |
| *** packet has joined #openstack-horizon | 00:18 | |
| *** cody-somerville has joined #openstack-horizon | 00:43 | |
| *** cody-somerville has quit IRC | 00:43 | |
| *** cody-somerville has joined #openstack-horizon | 00:43 | |
| *** cody-somerville has quit IRC | 00:44 | |
| *** ramishra has quit IRC | 00:51 | |
| *** ramishra has joined #openstack-horizon | 00:52 | |
| *** ramishra has quit IRC | 01:02 | |
| *** packet has quit IRC | 01:03 | |
| *** hightall has joined #openstack-horizon | 01:06 | |
| *** sunrenjie6 has joined #openstack-horizon | 01:09 | |
| *** r1chardj0n3s is now known as r1chardj0n3s_afk | 01:14 | |
| openstackgerrit | Cindy Lu proposed a change to openstack/horizon: add Column attribute for help tooltip in heading https://review.openstack.org/128804 | 01:19 |
|---|---|---|
| *** Longgeek has joined #openstack-horizon | 01:19 | |
| openstackgerrit | Cindy Lu proposed a change to openstack/horizon: add Column attribute for help tooltip in heading https://review.openstack.org/128804 | 01:21 |
| *** Longgeek has quit IRC | 01:25 | |
| *** bpokorny has quit IRC | 01:28 | |
| *** ramishra has joined #openstack-horizon | 01:32 | |
| *** vkmc has joined #openstack-horizon | 01:33 | |
| *** r1chardj0n3s_afk is now known as r1chardj0n3s | 01:38 | |
| *** clu_ has quit IRC | 01:39 | |
| *** ramishra has quit IRC | 01:42 | |
| *** ramishra has joined #openstack-horizon | 02:02 | |
| *** Longgeek has joined #openstack-horizon | 02:04 | |
| *** sunrenjie6 has quit IRC | 02:05 | |
| *** tqtran has quit IRC | 02:14 | |
| *** nlahouti has joined #openstack-horizon | 02:19 | |
| *** AliceR has quit IRC | 02:20 | |
| *** nlahouti has quit IRC | 02:23 | |
| *** jacalcat has joined #openstack-horizon | 02:25 | |
| *** lhcheng has quit IRC | 02:31 | |
| *** lhcheng has joined #openstack-horizon | 02:40 | |
| *** ramishra_ has joined #openstack-horizon | 02:58 | |
| *** ramishra has quit IRC | 03:00 | |
| *** ybathia has quit IRC | 03:07 | |
| *** ramishra_ has quit IRC | 03:10 | |
| *** jacalcat has quit IRC | 03:15 | |
| *** vkmc has quit IRC | 03:30 | |
| *** ramishra has joined #openstack-horizon | 03:32 | |
| *** harlowja is now known as harlowja_away | 03:32 | |
| *** harlowja_away is now known as harlowja | 03:34 | |
| openstackgerrit | A change was merged to openstack/horizon: _formset_table_row.html missing i18n tag. https://review.openstack.org/128614 | 03:38 |
| openstackgerrit | Zhenguo Niu proposed a change to openstack/horizon: Restrict the length of user name input https://review.openstack.org/128819 | 03:41 |
| *** vivekd has joined #openstack-horizon | 03:41 | |
| *** sbfox has joined #openstack-horizon | 03:46 | |
| *** masco has joined #openstack-horizon | 03:49 | |
| *** AliceR has joined #openstack-horizon | 03:58 | |
| *** amotoki__zzz has quit IRC | 04:00 | |
| *** sayan has joined #openstack-horizon | 04:12 | |
| *** vivekd has quit IRC | 04:17 | |
| *** lhcheng has quit IRC | 04:19 | |
| *** lhcheng has joined #openstack-horizon | 04:19 | |
| *** lhcheng has quit IRC | 04:24 | |
| *** ramishra has quit IRC | 04:24 | |
| *** vivekd has joined #openstack-horizon | 04:27 | |
| *** vivekd has left #openstack-horizon | 04:28 | |
| openstackgerrit | Travis Tripp proposed a change to openstack/horizon: Base Glance Metadata Definitions Admin UI https://review.openstack.org/104063 | 04:31 |
| *** nikunj2512 has joined #openstack-horizon | 04:43 | |
| *** hightall has quit IRC | 05:04 | |
| *** lhcheng has joined #openstack-horizon | 05:05 | |
| *** lhcheng has quit IRC | 05:08 | |
| *** lhcheng has joined #openstack-horizon | 05:09 | |
| *** hightall has joined #openstack-horizon | 05:12 | |
| *** masco has quit IRC | 05:17 | |
| *** ramishra has joined #openstack-horizon | 05:21 | |
| *** k4n0 has joined #openstack-horizon | 05:29 | |
| *** Longgeek has quit IRC | 05:35 | |
| *** HenryG has quit IRC | 05:36 | |
| *** sbfox has quit IRC | 05:37 | |
| *** harlowja is now known as harlowja_away | 05:41 | |
| *** sayan has quit IRC | 05:54 | |
| *** amotoki has joined #openstack-horizon | 05:56 | |
| *** coolsvap|afk is now known as coolsvap | 05:56 | |
| openstackgerrit | A change was merged to openstack/horizon: Hiding device name field when unhandled https://review.openstack.org/69423 | 05:57 |
| *** Longgeek has joined #openstack-horizon | 06:03 | |
| *** sayan has joined #openstack-horizon | 06:07 | |
| *** mrunge has joined #openstack-horizon | 06:08 | |
| *** masco has joined #openstack-horizon | 06:10 | |
| *** nlahouti has joined #openstack-horizon | 06:18 | |
| *** nlahouti has quit IRC | 06:18 | |
| *** nlahouti has joined #openstack-horizon | 06:19 | |
| *** johnma has quit IRC | 06:25 | |
| *** johnma has joined #openstack-horizon | 06:25 | |
| *** lhcheng has quit IRC | 06:29 | |
| *** sbfox has joined #openstack-horizon | 06:33 | |
| *** echevemaster has quit IRC | 06:36 | |
| *** dsneddon has quit IRC | 06:36 | |
| *** dsneddon has joined #openstack-horizon | 06:37 | |
| *** sayan has quit IRC | 06:37 | |
| *** lhcheng has joined #openstack-horizon | 06:45 | |
| *** exploreshaifali has joined #openstack-horizon | 06:47 | |
| *** nlahouti has quit IRC | 06:47 | |
| *** ramishra has quit IRC | 06:56 | |
| *** dsneddon has quit IRC | 06:57 | |
| *** nlahouti has joined #openstack-horizon | 06:57 | |
| *** ramishra has joined #openstack-horizon | 06:57 | |
| *** dsneddon has joined #openstack-horizon | 06:58 | |
| *** vokhrimenko has joined #openstack-horizon | 07:03 | |
| *** jtomasek has joined #openstack-horizon | 07:03 | |
| *** e0ne has joined #openstack-horizon | 07:05 | |
| *** e0ne has quit IRC | 07:07 | |
| *** sbfox has quit IRC | 07:09 | |
| *** e0ne has joined #openstack-horizon | 07:11 | |
| *** e0ne has quit IRC | 07:13 | |
| openstackgerrit | Matthias Runge proposed a change to openstack/horizon: Refactor metering dashboard https://review.openstack.org/107011 | 07:15 |
| *** jcoufal has joined #openstack-horizon | 07:16 | |
| *** ramishra has quit IRC | 07:16 | |
| *** ramishra has joined #openstack-horizon | 07:17 | |
| *** dsneddon has quit IRC | 07:18 | |
| *** dsneddon has joined #openstack-horizon | 07:18 | |
| *** nlahouti has quit IRC | 07:18 | |
| *** Ala has joined #openstack-horizon | 07:19 | |
| *** lhcheng has quit IRC | 07:20 | |
| *** HenryG has joined #openstack-horizon | 07:20 | |
| *** lhcheng has joined #openstack-horizon | 07:21 | |
| *** sayan has joined #openstack-horizon | 07:21 | |
| *** ramishra has quit IRC | 07:22 | |
| openstackgerrit | A change was merged to openstack/horizon: Fix an encoding exception when creating a public_url https://review.openstack.org/108450 | 07:22 |
| *** zz_ttrifonov is now known as ttrifonov | 07:23 | |
| *** lhcheng has quit IRC | 07:26 | |
| *** e0ne has joined #openstack-horizon | 07:28 | |
| *** cedrics has joined #openstack-horizon | 07:28 | |
| *** Ala has quit IRC | 07:30 | |
| *** cedrics has quit IRC | 07:31 | |
| *** e0ne has quit IRC | 07:32 | |
| *** cedrics has joined #openstack-horizon | 07:34 | |
| *** jprovazn has joined #openstack-horizon | 07:36 | |
| *** cedrics has quit IRC | 07:37 | |
| *** e0ne has joined #openstack-horizon | 07:37 | |
| *** cedrics has joined #openstack-horizon | 07:37 | |
| *** Ala has joined #openstack-horizon | 07:37 | |
| *** ramishra has joined #openstack-horizon | 07:38 | |
| openstackgerrit | Jiri Tomasek proposed a change to openstack/horizon: Restyled topbar to resemble UX guidelines https://review.openstack.org/117118 | 07:39 |
| *** cedricss has joined #openstack-horizon | 07:39 | |
| *** sunilm has joined #openstack-horizon | 07:39 | |
| *** cedrics_ has joined #openstack-horizon | 07:40 | |
| *** cedrics has quit IRC | 07:42 | |
| *** cedricss has quit IRC | 07:44 | |
| *** cedrics has joined #openstack-horizon | 07:44 | |
| *** cedricss has joined #openstack-horizon | 07:45 | |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/horizon: Imported Translations from Transifex https://review.openstack.org/128842 | 07:46 |
| *** cedrics_ has quit IRC | 07:46 | |
| *** cedrics has quit IRC | 07:49 | |
| *** e0ne has quit IRC | 07:49 | |
| *** e0ne has joined #openstack-horizon | 07:53 | |
| *** ramishra has quit IRC | 07:53 | |
| *** e0ne has quit IRC | 07:53 | |
| openstackgerrit | Akihiro Motoki proposed a change to openstack/horizon: Fix E128 errors in remaining openstack_dashboard/ https://review.openstack.org/125141 | 07:55 |
| openstackgerrit | Akihiro Motoki proposed a change to openstack/horizon: Fix E128 errors in openstack_dashboard/dashboards/project/ https://review.openstack.org/125140 | 07:55 |
| *** romainh has joined #openstack-horizon | 07:59 | |
| *** ygbo has joined #openstack-horizon | 08:04 | |
| *** ala_ has joined #openstack-horizon | 08:05 | |
| *** Ala has quit IRC | 08:08 | |
| *** jpich has joined #openstack-horizon | 08:16 | |
| *** MaxV has joined #openstack-horizon | 08:17 | |
| *** akrivoka has joined #openstack-horizon | 08:22 | |
| *** qba73 has joined #openstack-horizon | 08:27 | |
| mrunge | akrivoka, your patches for https://review.openstack.org/#/q/status:merged+project:openstack/horizon+branch:master+topic:bp/detail-pages-ia,n,z caused missed neutron mocks etc. Do you know, which commit fixed that? | 08:28 |
| akrivoka | mrunge: https://review.openstack.org/#/c/127237/ | 08:29 |
| jpich | mrunge: https://review.openstack.org/#/c/127237/ I think | 08:29 |
| mrunge | ah great, thanks | 08:29 |
| jpich | mrunge: Will we need that on stable/juno once it's cut? | 08:37 |
| *** hightall has quit IRC | 08:37 | |
| mrunge | jpich, I guess, no | 08:37 |
| *** hightall has joined #openstack-horizon | 08:37 | |
| jpich | mrunge: Ok! Thanks :) | 08:38 |
| mrunge | jpich, I was pulling in akrivoka s patches to https://github.com/redhat-openstack/horizon | 08:38 |
| mrunge | and unit tests were quite ugly then | 08:38 |
| jpich | mrunge: Got it! I get easily lost :-) | 08:39 |
| *** katyafervent_awa is now known as katyafervent | 08:41 | |
| *** ramishra has joined #openstack-horizon | 08:43 | |
| *** hightall has quit IRC | 08:48 | |
| *** AliceR has quit IRC | 08:51 | |
| *** e0ne has joined #openstack-horizon | 08:51 | |
| openstackgerrit | Matthias Runge proposed a change to openstack/horizon: Refactor metering dashboard https://review.openstack.org/107011 | 08:57 |
| *** ramishra has quit IRC | 09:01 | |
| *** exploreshaifali has quit IRC | 09:14 | |
| *** exploreshaifali has joined #openstack-horizon | 09:18 | |
| *** hhuang has joined #openstack-horizon | 09:25 | |
| *** exploreshaifali has quit IRC | 09:33 | |
| *** aix has joined #openstack-horizon | 09:38 | |
| *** hhuang has quit IRC | 09:42 | |
| openstackgerrit | Akihiro Motoki proposed a change to openstack/horizon: Do not log keystone token https://review.openstack.org/128859 | 09:43 |
| *** paulmillar has quit IRC | 09:55 | |
| openstackgerrit | A change was merged to openstack/horizon: Imported Translations from Transifex https://review.openstack.org/128842 | 09:56 |
| openstackgerrit | Vlad Okhrimenko proposed a change to openstack/horizon: Switch Region dropdown https://review.openstack.org/127544 | 09:58 |
| openstackgerrit | A change was merged to openstack/horizon: Make "VLAN", "VXLAN" and "GRE" translatable https://review.openstack.org/128732 | 09:59 |
| *** paulmillar has joined #openstack-horizon | 10:00 | |
| *** ZZelle has joined #openstack-horizon | 10:07 | |
| *** openstackgerrit has quit IRC | 10:19 | |
| *** openstackgerrit has joined #openstack-horizon | 10:20 | |
| *** akrivoka has quit IRC | 10:21 | |
| openstackgerrit | Swati Shukla proposed a change to openstack/horizon: Confirm Dialog for Delete operation https://review.openstack.org/128873 | 10:21 |
| *** sambetts has joined #openstack-horizon | 10:24 | |
| openstackgerrit | Marcin Karkocha proposed a change to openstack/horizon: Fix inconsistent displaying message in 'Other Field' in Update Metadata Form when you put something and delete it next. https://review.openstack.org/128878 | 10:27 |
| *** ramishra has joined #openstack-horizon | 10:28 | |
| *** ramishra has quit IRC | 10:32 | |
| *** cedrics has joined #openstack-horizon | 10:33 | |
| *** cedricss has quit IRC | 10:36 | |
| *** sayali has joined #openstack-horizon | 10:41 | |
| *** kmartin has quit IRC | 10:46 | |
| *** cedrics has quit IRC | 10:56 | |
| *** ramishra has joined #openstack-horizon | 10:57 | |
| *** akrivoka has joined #openstack-horizon | 10:59 | |
| *** nikunj2512 has quit IRC | 11:12 | |
| *** MaxV has quit IRC | 11:27 | |
| *** MaxV has joined #openstack-horizon | 11:28 | |
| *** MaxV has quit IRC | 11:32 | |
| mrunge | has anyone tested horizon on django-1.7? | 11:32 |
| mrunge | and if yes: do we have changes merged back? | 11:32 |
| mrunge | zigo? | 11:32 |
| amotoki | mrunge: zigo test django-1.7 and one unit test failure is still remaining. | 11:32 |
| mrunge | I saw some translation issues | 11:33 |
| amotoki | I have not tried django 1.7 yet.. | 11:34 |
| mrunge | amotoki, http://paste.fedoraproject.org/142396/13459234/ | 11:34 |
| mrunge | so, I wonder, how successful tests really were | 11:34 |
| *** ramishra has quit IRC | 11:35 | |
| zigo | mrunge: There seems to be another issue undetected with unit tests: http://lists.alioth.debian.org/pipermail/openstack-devel/2014-October/007488.html | 11:38 |
| zigo | I would need help to fix this ... | 11:38 |
| *** cedrics has joined #openstack-horizon | 11:41 | |
| *** MaxV has joined #openstack-horizon | 11:43 | |
| amotoki | i think so. it seems a different thing. | 11:45 |
| amotoki | I am not sure wsgi.py is called in unit tests. | 11:45 |
| *** julim has quit IRC | 11:46 | |
| openstackgerrit | A change was merged to openstack/horizon: Fix E128 errors in openstack_dashboard/dashboards/project/ https://review.openstack.org/125140 | 11:46 |
| tsufiev | does anyone know Lin Hua Cheng nick here? | 11:50 |
| mrunge | tsufiev, it's lcheng or lhcheng | 11:54 |
| mrunge | tsufiev, he's located in the US | 11:54 |
| mrunge | zigo, it looks like your issue is the same as mine with Django-1.7 | 11:56 |
| zigo | mrunge: Ok, good to know. Did you file a bug? | 11:56 |
| mrunge | well, same issue, different location | 11:56 |
| mrunge | zigo, not yet. did you? | 11:57 |
| zigo | mrunge: No, please do ! :) | 11:57 |
| mrunge | zigo, this is Icehouse version in Debian, right? | 12:00 |
| zigo | mrunge: There's Icehouse in Sid/Jessie, and Juno in Experimental. | 12:00 |
| zigo | I need both to work with Django 1.7. :/ | 12:00 |
| mrunge | no. | 12:01 |
| mrunge | really? | 12:01 |
| mrunge | so, this is critical, right? | 12:02 |
| mrunge | zigo; https://bugs.launchpad.net/horizon/+bug/1382023 | 12:02 |
| *** radez_g0n3 is now known as radez | 12:04 | |
| zigo | mrunge: yes, this is critical, and I would very much prefer it to be fixed before Jessie is frozen on the 5th of November. | 12:06 |
| mrunge | zigo, the question is more: is this so critical to prevent a Juno release? | 12:06 |
| zigo | mrunge: As always, people in OpenStack will say "but we only declared compatibility with an older version, so we don't care" ... | 12:07 |
| zigo | Reality, people try to do their best. | 12:07 |
| zigo | But that can't block a release. | 12:07 |
| *** rbertram has quit IRC | 12:14 | |
| *** rbertram has joined #openstack-horizon | 12:14 | |
| zigo | mrunge: Also, as long as we have a patch, it can be included in the distro, even if it's not merged upstream yet. | 12:15 |
| zigo | So the important bit is to have someone to work on it... | 12:15 |
| mrunge | zigo, I would want a patch to be merged... | 12:15 |
| zigo | mrunge: Of course, that's *better*, but not mandatory for me. | 12:16 |
| *** ramishra has joined #openstack-horizon | 12:16 | |
| zigo | As long as I have a patch to use... | 12:16 |
| mrunge | zigo, are you applying some other patches on horizon? | 12:19 |
| zigo | mrunge: Yes, I do. | 12:19 |
| mrunge | ... is there a patch to upstream? | 12:19 |
| zigo | fix-dashboard-django-wsgi.patch | 12:20 |
| zigo | fix-dashboard-manage.patch | 12:20 |
| zigo | fixed-horizon-MANIFEST.in.patch | 12:20 |
| zigo | disable-failed-django-1.7-test.patch | 12:20 |
| zigo | Remove_selenium_dependency_when_not_using_selenium_tests.patch | 12:20 |
| zigo | That's what I currently have. | 12:20 |
| mrunge | zigo, could you submit them for review? | 12:20 |
| mrunge | if manageable, I mean | 12:20 |
| *** vkmc has joined #openstack-horizon | 12:21 | |
| mrunge | it would be a shame, not to submit patches back upstream | 12:21 |
| mrunge | ^ zigo | 12:21 |
| *** vkmc has quit IRC | 12:21 | |
| *** vkmc has joined #openstack-horizon | 12:21 | |
| zigo | mrunge: Some don't need to, and are Debian specific, some I already tried and failed to gather upstream attention or got negative reviews, and the last one is from upstream but not yet merged (probably it is right now...). | 12:21 |
| zigo | mrunge: Most Django 1.7 patches, I was the one to send them for review. | 12:22 |
| zigo | Thanks to Raphael Hertzhog for his help on them though ! :) | 12:22 |
| mrunge | zigo, could you probably link them to the django-1.7 bug report? | 12:22 |
| mrunge | .. just to make them more prominent? | 12:22 |
| zigo | mrunge: Link the patches which were aready merged? | 12:23 |
| zigo | I don't think I'd have time to do that right now. | 12:23 |
| * zigo is busy with building and uploading Juno final tags to Debian Experimental. | 12:23 | |
| mrunge | no, I meant to link unmerged patches | 12:24 |
| *** hightall has joined #openstack-horizon | 12:26 | |
| jpich | mrunge, zigo: Updating the wsgi file like in https://www.pythonanywhere.com/forums/topic/1629/#id_post_11030 appears to resolve the AppRegistryNotReady issue for me (I haven't checked yet if it's backwards compatible though...) | 12:27 |
| zigo | mrunge: All Django 1.7 patches I sent were merged, at least for Juno, not all were backported to Icehouse though. | 12:27 |
| mrunge | zigo, thank you, good to hear | 12:27 |
| zigo | jpich: Oh, thanks a lot Julie !!! :) | 12:27 |
| * zigo reads | 12:27 | |
| *** ramishra has quit IRC | 12:28 | |
| zigo | jpich: So, if I follow what you send, we should do: | 12:29 |
| zigo | from django.core.wsgi import get_wsgi_application | 12:29 |
| zigo | application = get_wsgi_application() | 12:29 |
| zigo | jpich: But where should that be replaced? | 12:29 |
| jpich | Instead of import django.core.handlers.wsgi and application = django.core.handlers.wsgi.WSGIHandler() yeah, in https://github.com/openstack/horizon/blob/master/openstack_dashboard/wsgi/django.wsgi | 12:29 |
| mrunge | jpich, thanks, that looks promising | 12:30 |
| * zigo currently installs Jessie to test this out. | 12:30 | |
| openstackgerrit | Vlad Okhrimenko proposed a change to openstack/horizon: Hide the "Host Aggregates" admin panel if not enabled https://review.openstack.org/121127 | 12:31 |
| *** absubram has quit IRC | 12:31 | |
| mrunge | jpich, apparently, that works in Django-1.6 as well | 12:31 |
| mrunge | Django-1.6 and icehouse I meant | 12:31 |
| jpich | Sounds good! Let's see if we can support back to 1.4 with this... | 12:32 |
| mrunge | jpich, are you going to submit a patch? | 12:32 |
| jpich | mrunge: For now I'm updating the bug and figuring out the backwards compatibility implications | 12:32 |
| mrunge | jpich, thank you very much, awesome! | 12:32 |
| *** neelashah has joined #openstack-horizon | 12:32 | |
| jpich | Though I guess we could make a conditional import based on django version if it comes down to it probably | 12:33 |
| openstackgerrit | Adrien Vergé proposed a change to openstack/horizon: [Sahara] Support Cinder availability zones https://review.openstack.org/128906 | 12:33 |
| zigo | jpich: How would you test Django's version? Is there something smarter than just pkg-resources ? | 12:34 |
| jpich | zigo: There's a django_version variable of some sorts, we use it in a few places, it's pretty hackish. But it looks we won't need to here, woot :) | 12:35 |
| jpich | I'll submit a patch then | 12:35 |
| amotoki | yeah! hacking patch series is completing... the final one goes into the gate. | 12:36 |
| jpich | all the progress :-) | 12:38 |
| mrunge | jpich, would be nice, if this patch could make it into Juno | 12:40 |
| jpich | mrunge: Juno's already been tagged so it's not going to happen for the release today, but possibly there is an argument for the first stable point release | 12:40 |
| jpich | though it's a bit featurish? | 12:41 |
| mrunge | well, if it works from Django-1.4 and up? | 12:41 |
| jpich | If the patches are minimal, it certainly sounds reasonable to me | 12:42 |
| jpich | you're the stable maintainer either way ;) | 12:42 |
| mrunge | who? | 12:42 |
| mrunge | ;-) | 12:42 |
| *** johnma has quit IRC | 12:45 | |
| *** _crobertsrh is now known as crobertsrh | 12:52 | |
| *** ramishra has joined #openstack-horizon | 12:54 | |
| *** shadower39 has joined #openstack-horizon | 12:57 | |
| openstackgerrit | Julie Pichon proposed a change to openstack/horizon: Update WSGI app creation to be compatible with Django 1.7 https://review.openstack.org/128911 | 13:00 |
| *** shadower39 has quit IRC | 13:01 | |
| *** shadower78 has joined #openstack-horizon | 13:02 | |
| *** shadower has quit IRC | 13:03 | |
| *** shadower78 is now known as shadower | 13:03 | |
| openstackgerrit | Jiri Tomasek proposed a change to openstack/horizon: Restyled sidebar to resemble UX guidelines https://review.openstack.org/126289 | 13:03 |
| *** tnovacik has joined #openstack-horizon | 13:05 | |
| *** neelashah has quit IRC | 13:06 | |
| *** tzumainn has joined #openstack-horizon | 13:06 | |
| *** johnma has joined #openstack-horizon | 13:12 | |
| *** thiagop has joined #openstack-horizon | 13:14 | |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/horizon: Merge tag '2014.2' https://review.openstack.org/128916 | 13:20 |
| *** ramishra has quit IRC | 13:20 | |
| *** ramishra has joined #openstack-horizon | 13:21 | |
| *** julim has joined #openstack-horizon | 13:21 | |
| openstackgerrit | Marcin Karkocha proposed a change to openstack/horizon: Fix misleading message in Update Metadata Form. First you put some data to Other field. In next step you delete it. Then you give message 'invalid key name'. This is inconsistent to this field behavior. https://review.openstack.org/128878 | 13:23 |
| *** peristeri has joined #openstack-horizon | 13:28 | |
| *** johnma has quit IRC | 13:29 | |
| *** jacalcat has joined #openstack-horizon | 13:31 | |
| openstackgerrit | A change was merged to openstack/horizon: Fix E128 errors in remaining openstack_dashboard/ https://review.openstack.org/125141 | 13:32 |
| *** absubram has joined #openstack-horizon | 13:33 | |
| thiagop | tsufiev: ping | 13:35 |
| *** jprovazn has quit IRC | 13:37 | |
| *** hightall has quit IRC | 13:38 | |
| jpich | amotoki: Thanks for linking to the 1.7 release notes in the wsgi bug, that was useful information. I need to read them more carefully next time rather than skimming :) | 13:47 |
| tsufiev | thiagop, pong | 13:48 |
| *** shufflebot has quit IRC | 13:48 | |
| thiagop | henrynash: tsufiev Hi! I have ended a PoC of that 2nd part of RBAC on Horizon (the one that retrieves the policy from the endpoint on keystone) | 13:48 |
| thiagop | tsufiev: would you like to take a look? | 13:49 |
| *** cody-somerville has joined #openstack-horizon | 13:50 | |
| *** cody-somerville has quit IRC | 13:50 | |
| *** cody-somerville has joined #openstack-horizon | 13:50 | |
| *** hightall has joined #openstack-horizon | 13:51 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 13:56 | |
| *** sigmavirus24 has left #openstack-horizon | 13:59 | |
| *** hightall has quit IRC | 14:00 | |
| *** ramishra has quit IRC | 14:00 | |
| *** doug-fish has joined #openstack-horizon | 14:00 | |
| *** hightall has joined #openstack-horizon | 14:01 | |
| *** ericpeterson has joined #openstack-horizon | 14:02 | |
| ericpeterson | hey all, checking in if there have been any fixes around the project picker in the last 18 hours or so? seeing problems on master | 14:03 |
| ericpeterson | when you switch projects, the old token is purged / deactived..... and the session does not get updated with the new token | 14:03 |
| mrunge | ericpeterson, if I'm not totally wrong, there was a change on that topic | 14:05 |
| ericpeterson | that would be awesome if a fix came in recently | 14:08 |
| ericpeterson | seems like yesterday things were pretty broken | 14:08 |
| ericpeterson | (could be yesterday - 1 day or so) | 14:09 |
| *** ramishra has joined #openstack-horizon | 14:09 | |
| thiagop | ericpeterson: I found what caused it | 14:10 |
| thiagop | ericpeterson: just couldn't find the main cause | 14:10 |
| ericpeterson | yeah, the user tries to get updated, but it doesn't stick | 14:10 |
| jpich | Is that a problem on Juno too or is it new? | 14:11 |
| ericpeterson | I think it's been around for a few days at least, maybe longer | 14:11 |
| *** cedricss has joined #openstack-horizon | 14:11 | |
| ericpeterson | request._cached_user = user | 14:11 |
| ericpeterson | request.user = user | 14:11 |
| ericpeterson | + request.session.modified = True | 14:11 |
| ericpeterson | I tried this in DOA ^ | 14:12 |
| thiagop | it was caused by this commit | 14:12 |
| thiagop | https://github.com/openstack/horizon/commit/16db58fabad8934b8fbdfc6aee0361cc138b20af | 14:12 |
| thiagop | when you remove the decorator from the can_access function, it works like a charm | 14:12 |
| tsufiev | thiagop, sure, this would be very interesting | 14:12 |
| *** jprovazn has joined #openstack-horizon | 14:12 | |
| thiagop | it appears that the context being received by the decorator remains the same of before changing the project | 14:13 |
| thiagop | (I have printed the token before and after project changes, with and without the decorator) | 14:13 |
| ericpeterson | excellent find btw ;) | 14:14 |
| *** sunilm has quit IRC | 14:14 | |
| *** cedrics has quit IRC | 14:14 | |
| *** amotoki_ has joined #openstack-horizon | 14:14 | |
| thiagop | when you get the decorator out, the context on "can_acess" is correct (to the new project) | 14:15 |
| thiagop | it's VERY weird... I struggled with it for the whole day yesterday | 14:15 |
| ericpeterson | it's like there is a copy of the session, and we modify one version, and then the old version gets saved last | 14:15 |
| *** david-lyle has joined #openstack-horizon | 14:16 | |
| tsufiev | thiagop, ericpeterson: shouldn't this problem be handled in https://github.com/openstack/horizon/commit/16db58fabad8934b8fbdfc6aee0361cc138b20af#diff-3ff8a61f264181ecfecf2ba85e7781c6R63 ? | 14:17 |
| *** amerine_ has joined #openstack-horizon | 14:17 | |
| tsufiev | i thought that one project can be distinguished from another by their tokens | 14:17 |
| *** amerine has quit IRC | 14:17 | |
| thiagop | tsufiev: You're right, but the context being received in this function is, for some reason, the old context (before the new token being inserted in the session) | 14:19 |
| ericpeterson | the tokens and the projects are a bit differet..... a token will either give access to a project or not | 14:19 |
| ericpeterson | yeah, we have the old token, get a new token..... we then invalidate the old token | 14:20 |
| thiagop | tsufiev: about the PoC, #links: https://etherpad.openstack.org/p/poc-endpoint-policy-horizon and https://github.com/tpborion/horizon/compare/poc-endpoint-policy?expand=1 | 14:20 |
| ericpeterson | and we try to save the new token... | 14:20 |
| ericpeterson | but we reuse the old token | 14:20 |
| ericpeterson | and then all the things fail in horrific fashion | 14:20 |
| tsufiev | thiagop, thanks, will take a look! | 14:21 |
| ericpeterson | is there a bug open on this? | 14:22 |
| ericpeterson | cuz it's like super bad | 14:22 |
| *** packet has joined #openstack-horizon | 14:22 | |
| thiagop | we can file one now :) | 14:22 |
| *** athomas has quit IRC | 14:23 | |
| ericpeterson | if you can't change projects, that seems like a complete show stopper | 14:23 |
| thiagop | I was mainly preoccupied with the solution | 14:23 |
| ericpeterson | yeah, me too. you want to file or I? | 14:23 |
| thiagop | I can do that | 14:23 |
| ericpeterson | thanks! :D | 14:23 |
| *** vijendar has quit IRC | 14:26 | |
| *** nikunj2512 has joined #openstack-horizon | 14:26 | |
| ericpeterson | why is this stuff being saved on the session anyways? seems like allowed stuff should not be on the session | 14:26 |
| ericpeterson | a memoized version maybe, but on the session is not right | 14:27 |
| jpich | I guess because we don't really have a way to store anything horizon specific so that's where stuff tends to go | 14:27 |
| ericpeterson | memoize works too, we use it elsewhere | 14:27 |
| ericpeterson | hindsight is 20:20.... I'm just blathering on | 14:28 |
| tsufiev | ericpeterson, guess that's because we need same behavior across several apache web-servers on many controller nodes and cookie-based sessions provide that? | 14:29 |
| *** johnma has joined #openstack-horizon | 14:29 | |
| *** athomas has joined #openstack-horizon | 14:30 | |
| ericpeterson | if you are rellying upon cookie based sessions, that is a non starter | 14:30 |
| ericpeterson | cookie based sessions are only usefull in very small deployments like devstack | 14:30 |
| ericpeterson | with the cookie size maxing at 4k and our tokens being huge | 14:31 |
| thiagop | ericpeterson: https://bugs.launchpad.net/horizon/+bug/1382079 | 14:31 |
| ericpeterson | thiagop thanks, I think I can see a way to fix, trying to understand orig patch | 14:32 |
| ericpeterson | we want to memoize this, and not use sessions | 14:32 |
| *** exploreshaifali has joined #openstack-horizon | 14:32 | |
| thiagop | ericpeterson: sorry, I didn't get your point. Can you elaborate? :) | 14:34 |
| ericpeterson | here is an example of all the new items we are keeping on the session (appologies ahead of time): | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.project.overview.panel.Overview | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.project.instances.panel.Instances | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.project.volumes.panel.Volumes | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.project.images.panel.Images | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.project.access_and_security.panel.AccessAndSecurity | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.project.network_topology.panel.NetworkTopology | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.project.networks.panel.Networks | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.project.routers.panel.Routers | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.project.loadbalancers.panel.LoadBalancer | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.project.firewalls.panel.Firewall | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.project.vpn.panel.VPN | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.project.containers.panel.Containers | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.project.stacks.panel.Stacks | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.project.databases.panel.Databases | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.project.database_backups.panel.Backups | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.project.data_processing.clusters.panel.ClustersPanel | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.project.data_processing.cluster_templates.panel.ClusterTemplatesPanel | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.project.data_processing.nodegroup_templates.panel.NodegroupTemplatesPanel | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.project.data_processing.job_executions.panel.JobExecutionsPanel | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.project.data_processing.jobs.panel.JobsPanel | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.project.data_processing.job_binaries.panel.JobBinariesPanel | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.project.data_processing.data_sources.panel.DataSourcesPanel | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.project.data_processing.data_image_registry.panel.ImageRegistryPanel | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.project.data_processing.data_plugins.panel.PluginsPanel | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.project.dashboard.Project | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.admin.overview.panel.Overview | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.admin.metering.panel.Metering | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.admin.hypervisors.panel.Hypervisors | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.admin.aggregates.panel.Aggregates | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.admin.instances.panel.Instances | 14:35 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.admin.volumes.panel.Volumes | 14:36 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.admin.flavors.panel.Flavors | 14:36 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.admin.images.panel.Images | 14:36 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.admin.networks.panel.Networks | 14:36 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.admin.routers.panel.Routers | 14:36 |
| jpich | This is pretty horrifying. Please use paste.openstack.org for that kind of long paste in the future? You're probably getting throttled for spam too :-) | 14:36 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.admin.defaults.panel.Defaults | 14:36 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.admin.info.panel.Info | 14:36 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.admin.dashboard.Admin | 14:36 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.identity.projects.panel.Tenants | 14:36 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.identity.users.panel.Users | 14:36 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.identity.dashboard.Identity | 14:36 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.settings.user.panel.UserPanel | 14:36 |
| ericpeterson | being called for key = openstack_dashboard.dashboards.settings.password.panel.PasswordPanel | 14:36 |
| ericpeterson | and each one of those gets a func kept with it | 14:36 |
| ericpeterson | thiagop there is a memoize decorator you can add to cache a method | 14:36 |
| *** coolsvap is now known as coolsvap|afk | 14:36 | |
| vkmc | ericpeterson, please use http://paste.openstack.org | 14:37 |
| *** jacalcat1 has joined #openstack-horizon | 14:37 | |
| ericpeterson | thanks, sorry about that | 14:37 |
| *** jacalcat has quit IRC | 14:37 | |
| thiagop | ericpeterson: got it | 14:37 |
| ericpeterson | https://github.com/openstack/horizon/blob/master/horizon/utils/memoized.py has the function that would work mo betta I think | 14:38 |
| ericpeterson | decorator / whatever | 14:38 |
| tsufiev | ericpeterson, do you think we're hitting 4K limit here? | 14:38 |
| jpich | It seems the goal is different though, since memoized only works once per request? | 14:38 |
| jpich | while I assume the other one was caching it for the session length | 14:38 |
| ericpeterson | which is exactly what I think you'd want, for allowed checks | 14:39 |
| jpich | In case someone gets their access rights downgraded while doing naughty things? | 14:40 |
| ericpeterson | and the other once makes 43 session modifications per reques | 14:40 |
| ericpeterson | t | 14:40 |
| *** k4n0 has quit IRC | 14:40 | |
| zigo | jpich: Thanks for your WGSI application patch. | 14:41 |
| ericpeterson | sorry, 43 session writes initially on login | 14:41 |
| jpich | zigo: Cool! So it does resolve the problem for you too? | 14:41 |
| zigo | jpich: Can't tell yet, still reinstalling my test vm... | 14:42 |
| zigo | jpich: I'll comment in the review. | 14:42 |
| jpich | ericpeterson: A once-off cost at login may be more acceptable than when loading a page each time though, depending on the impact... | 14:42 |
| jpich | zigo: I'm fairly confident/hopeful this is the same issue but it's always good to have confirmation so, thanks! Let me know. | 14:43 |
| ericpeterson | yeah, I am floundering through geting this fixed, a bit all over the place atm | 14:43 |
| *** cody-somerville has quit IRC | 14:44 | |
| jpich | Ok, don't let me interrupt your thinking aloud then :-) Good luck | 14:45 |
| *** akrivoka has quit IRC | 14:47 | |
| *** openstackgerrit has quit IRC | 14:48 | |
| *** openstackgerrit has joined #openstack-horizon | 14:50 | |
| *** akrivoka has joined #openstack-horizon | 14:50 | |
| *** hightall has quit IRC | 14:53 | |
| tzumainn | hiya - does anyone know who does the horizon pypi releases? we'd like to do one for tuskar-ui, but I'm not entirely sure what the requirements for that are | 14:55 |
| jpich | tzumainn: I *think* openstack projects get released onto pypi automatically whenever a tag is pushed | 14:56 |
| *** vijendar has joined #openstack-horizon | 14:56 | |
| thiagop | ericpeterson: I'd like very much to know why these contexts are coming differently | 14:56 |
| jpich | tzumainn: Do you mean you need today's release for tuskar-ui, or you want to have tuskar-ui released there too? | 14:56 |
| tzumainn | jpich, oh, I think I mean I just need to know how to push a tag : ) | 14:57 |
| * thiagop going lunch! | 14:57 | |
| openstackgerrit | Marcin Karkocha proposed a change to openstack/horizon: Fix misleading message in Update Metadata Form. https://review.openstack.org/128878 | 14:57 |
| tzumainn | jpich, I think I need permissions somewhere somehow, not sure if I can get those permissions or if I need to ask someone else to create the tag for me. . . ? | 14:58 |
| jpich | tzumainn: I think anyone on the core team can do that? Or maybe only PTLs? I would suggest asking the way more knowledgeable #openstack-infra folks :-) They should be able to let you know if the repo is set up for it or not, too | 14:58 |
| david-lyle | tzumainn, I think that's me | 14:58 |
| david-lyle | are you wanting tuskar-ui tagged? | 14:59 |
| tzumainn | david-lyle, ah, yep! | 14:59 |
| david-lyle | what tag? | 14:59 |
| *** ramishra has quit IRC | 15:00 | |
| tzumainn | 0.2.0 | 15:00 |
| *** KurtMartin has joined #openstack-horizon | 15:00 | |
| david-lyle | everything's in that needs to be? | 15:00 |
| tzumainn | yep! | 15:00 |
| *** jprovazn is now known as jprovazn_afk | 15:00 | |
| tzumainn | jpich, thanks for the info btw! | 15:03 |
| *** MaxV has quit IRC | 15:03 | |
| *** hightall has joined #openstack-horizon | 15:05 | |
| jpich | tzumainn: You're quite welcome | 15:06 |
| openstackgerrit | A change was merged to openstack/horizon: Update WSGI app creation to be compatible with Django 1.7 https://review.openstack.org/128911 | 15:07 |
| ericpeterson | http://paste.openstack.org/show/121487/ seems to be a version of a fix thiagop | 15:08 |
| ericpeterson | not sure that is really correct tho | 15:08 |
| ericpeterson | it IS better, tho | 15:08 |
| ericpeterson | or less broken, at least | 15:08 |
| *** ramishra has joined #openstack-horizon | 15:10 | |
| *** masco has quit IRC | 15:16 | |
| *** nikunj2512 has quit IRC | 15:18 | |
| david-lyle | tzumainn, tag pushed and release uploaded | 15:20 |
| tzumainn | david-lyle, thanks! | 15:20 |
| tzumainn | david-lyle, so is the correct procedure to ping you when we need a release? or is there another preferable method? | 15:21 |
| *** ramishra has quit IRC | 15:22 | |
| david-lyle | tzumainn, ping me is the route, there is a ptl group for horizon that can push tags | 15:22 |
| *** nikunj2512 has joined #openstack-horizon | 15:22 | |
| tzumainn | david-lyle, okay, thanks again! | 15:23 |
| david-lyle | np | 15:23 |
| *** ala_ has quit IRC | 15:24 | |
| *** rebelagentm has joined #openstack-horizon | 15:25 | |
| *** vijendar has quit IRC | 15:28 | |
| *** nikunj2512 has quit IRC | 15:36 | |
| *** qba73 has quit IRC | 15:37 | |
| *** qba73 has joined #openstack-horizon | 15:38 | |
| openstackgerrit | Abishek Subramanian proposed a change to openstack/horizon: Display profile information in network details https://review.openstack.org/76787 | 15:41 |
| *** sayan is now known as sayan|afk | 15:42 | |
| *** hightall has quit IRC | 15:46 | |
| openstackgerrit | Mingyan Bao proposed a change to openstack/horizon: Add pagination support to the volume page https://review.openstack.org/118450 | 15:50 |
| *** exploreshaifali has quit IRC | 15:54 | |
| *** david-lyle has quit IRC | 15:54 | |
| *** qba73 has quit IRC | 15:54 | |
| ericpeterson | thiagop: found the issue / fix | 15:55 |
| ericpeterson | the issue we both see is related to sessions going over the limit with cookies | 15:57 |
| ericpeterson | if you hook up mysql as your session backend, stuff works then | 15:57 |
| ericpeterson | or if you take that pastebin change, that would work too | 15:57 |
| ericpeterson | or if you have a service catalog with about 5 bytes, that would help too ;) | 15:58 |
| *** cedricss has quit IRC | 15:59 | |
| jpich | Maybe we could disintegrate projects to make sure the cookie doesn't get too large ever | 16:04 |
| *** jacalcat1 has quit IRC | 16:04 | |
| *** gugl2 has quit IRC | 16:05 | |
| *** aix has quit IRC | 16:07 | |
| tzumainn | huh | 16:09 |
| tzumainn | now I want to eat cookies | 16:09 |
| *** lhcheng has joined #openstack-horizon | 16:10 | |
| *** Longgeek has quit IRC | 16:12 | |
| *** ygbo has quit IRC | 16:12 | |
| openstackgerrit | Paul Karikh proposed a change to openstack/horizon: Remove admin role name 'admin' hardcode https://review.openstack.org/123741 | 16:15 |
| *** KurtMartin has quit IRC | 16:15 | |
| *** KurtMartin has joined #openstack-horizon | 16:16 | |
| *** vokhrimenko has quit IRC | 16:18 | |
| *** AliceR has joined #openstack-horizon | 16:18 | |
| *** akrivoka has quit IRC | 16:19 | |
| ericpeterson | lhcheng - do you know about memoize per session? | 16:21 |
| ericpeterson | seems like we'd want something like that for this case maybe | 16:21 |
| *** david-lyle has joined #openstack-horizon | 16:21 | |
| * thiagop I'm back | 16:22 | |
| *** ramishra has joined #openstack-horizon | 16:22 | |
| ericpeterson | thiagop: got a fix | 16:22 |
| ericpeterson | kind of | 16:22 |
| ericpeterson | the problem is that we are overflowing the session with this change | 16:22 |
| ericpeterson | if you used mysql sessions, it's fine | 16:22 |
| ericpeterson | or if you just memoize it that works too | 16:22 |
| *** vijendar has joined #openstack-horizon | 16:23 | |
| ericpeterson | http://paste.openstack.org/show/121487/ thiagop is the memoized version of that | 16:26 |
| *** jcoufal has quit IRC | 16:26 | |
| thiagop | ericpeterson: I've not tryied it yet, but maybe we can remove this decorator and use just memoize? What do you think? | 16:28 |
| thiagop | (on "allowed") | 16:28 |
| *** lsmola has quit IRC | 16:29 | |
| david-lyle | regardless, that info should not be shoved into the session data | 16:30 |
| lhcheng | ericpeterson: don't recall we have a code that memoize by session. closest is probably memoizing it by token | 16:32 |
| *** sigmavirus24 has joined #openstack-horizon | 16:34 | |
| *** ramishra has quit IRC | 16:34 | |
| zigo | jpich: Your wsgi application patch fixes the issue for me in Icehouse! :) | 16:36 |
| jpich | zigo: \o/ | 16:36 |
| zigo | jpich: I didn't test Juno yet, but I'm quite sure it's going to be the same. | 16:36 |
| david-lyle | lhcheng, how are we memoizing by session? I thought it was scoped to the request/response cycle | 16:37 |
| zigo | jpich: Or is it merged in Juno already? | 16:37 |
| zigo | Ah no... | 16:37 |
| zigo | Sorry! :P | 16:37 |
| zigo | jpich: Please do backport it to both Icehouse and Juno. | 16:38 |
| jpich | zigo: I would assume so yeah. The master patch already merged, I'll submit a juno backport when the stable branch is properly open | 16:39 |
| jpich | yup! | 16:39 |
| *** jpich has quit IRC | 16:41 | |
| *** sayan has joined #openstack-horizon | 16:43 | |
| *** lsmola has joined #openstack-horizon | 16:44 | |
| *** sbfox has joined #openstack-horizon | 16:45 | |
| lhcheng | david-lyle, we don't have anything memoizing by session :) Closest I could think of memoizing up to scope of token. token != session, user can switch tokens in a single session. | 16:47 |
| david-lyle | not sure how it's even across token? | 16:48 |
| lhcheng | david-lyle, we have a function that is memoized by keyword argument, we passed the token as the key to used to store in cache. | 16:49 |
| *** jamespage_ has joined #openstack-horizon | 16:49 | |
| lhcheng | david-lyle, we used it in caching get_project_list() | 16:50 |
| david-lyle | but we're not using a cache backend, how is this info not getting cleaned up when response is sent? | 16:51 |
| *** jamespage_ has quit IRC | 16:51 | |
| david-lyle | ah, because we're implementing our own memoize for some reason | 16:52 |
| lhcheng | stored in memory, it gets cleaned up when the user logs out | 16:52 |
| david-lyle | was referencing any of the other 18 implemenations of memoize in python and django | 16:53 |
| david-lyle | in fact memoize is built in to django | 16:53 |
| david-lyle | and is starting a deprecation cycle in 1.7 | 16:53 |
| david-lyle | why we built our own ?? who knows | 16:54 |
| *** coolsvap|afk is now known as coolsvap | 16:55 | |
| thiagop | david-lyle lhcheng so, you think we should not be using memoized? | 16:56 |
| david-lyle | memoize is fine | 16:58 |
| david-lyle | just odd to me that we reimplemented it | 16:58 |
| david-lyle | may want to look at replacing with the django implementation, although that seems to be transitioning | 16:59 |
| thiagop | david-lyle: even with the imminent deprecation of the django's impl? | 16:59 |
| ericpeterson | let's write a new version of memoize for this case :P | 16:59 |
| david-lyle | we have some strange issues with the weak_ref in our implementation | 17:00 |
| * ericpeterson ducks | 17:00 | |
| david-lyle | thiagop, it's transitioning | 17:00 |
| *** ybathia has joined #openstack-horizon | 17:00 | |
| david-lyle | http://django.readthedocs.org/en/latest/releases/1.7.html#django-utils-functional-memoize-function | 17:01 |
| lhcheng | david-lyle: I joined late, I was just responding to ericpeterson if we have memoized by session. didn't have a context of the question if we should memoize or not | 17:01 |
| *** sbfox has quit IRC | 17:02 | |
| david-lyle | lhcheng, the recent fix to can_access stores a ton of stuff on the session that is overflowing the session cookie | 17:02 |
| *** sbfox has joined #openstack-horizon | 17:02 | |
| david-lyle | question is can we move that to memoize | 17:02 |
| lhcheng | david-lyle I've looked at memoized in django, it is fairly primitive. It takes the a number of the args to used as key, there is no guarantee it'll use token as the key to the cache. | 17:03 |
| david-lyle | I think we can, token based would actually be perfect, because the results can change on token rescope | 17:03 |
| david-lyle | lhcheng, was hoping it was more mature | 17:03 |
| david-lyle | :( | 17:03 |
| david-lyle | that's disappointing | 17:04 |
| david-lyle | there are whole pypi projects implementing memoize in django as well | 17:04 |
| ericpeterson | http://paste.openstack.org/show/121487/ lhcheng was the change around current issue we see | 17:04 |
| ericpeterson | you can see we stored a ton of stuff on the session with a recent change | 17:05 |
| *** ArcTanSusan has joined #openstack-horizon | 17:05 | |
| lhcheng | recent merge? | 17:06 |
| lhcheng | haven't seen that before... | 17:06 |
| thiagop | lhcheng: https://github.com/openstack/horizon/commit/16db58fabad8934b8fbdfc6aee0361cc138b20af | 17:08 |
| *** sbfox has quit IRC | 17:09 | |
| lhcheng | david-lyle, I agree. We should not putting caching those information in the session.. | 17:11 |
| david-lyle | we're already pushing the limit without it | 17:11 |
| lhcheng | we know better than adding more data to session :P | 17:12 |
| david-lyle | indeed we do | 17:12 |
| lhcheng | the api calls can be reduced by just memoizing on the service calls | 17:13 |
| david-lyle | as long as that works with mox, which was the main reason for the change | 17:13 |
| lhcheng | the service made on the panels are just checked if the extension is enabled. | 17:13 |
| lhcheng | we could just memoized it for the whole app | 17:13 |
| david-lyle | different neutron endpoints could potentially have different extension enabled, no? | 17:14 |
| lhcheng | there was a patch the works around that.. | 17:14 |
| lhcheng | good call about neutron endpoint | 17:14 |
| david-lyle | isn't that patch what merged? | 17:14 |
| lhcheng | david-lyle checking... | 17:16 |
| david-lyle | pretty sure it is | 17:17 |
| david-lyle | just not mentioned in the bug report | 17:17 |
| *** harlowja_away is now known as harlowja | 17:19 | |
| lhcheng | david-lyle https://review.openstack.org/#/c/121127/ | 17:20 |
| *** johnma has quit IRC | 17:20 | |
| *** sbfox has joined #openstack-horizon | 17:21 | |
| lhcheng | this will need to be rebased from the commit that thiagop gave me | 17:21 |
| *** sbfox has quit IRC | 17:22 | |
| david-lyle | wonder if we should just revert the panel cache | 17:22 |
| david-lyle | seems misguided | 17:22 |
| david-lyle | that's causing the current issues | 17:23 |
| david-lyle | session scope is too long | 17:23 |
| thiagop | david-lyle: +1, at least while we can't think of a better solution | 17:23 |
| lhcheng | david-lyle, I agree | 17:23 |
| thiagop | ericpeterson: the solution using memoized seems to be always calling the policy validation | 17:24 |
| lhcheng | ericpeterson, what that the cause of your issue yesterday? | 17:24 |
| thiagop | It's like it wasn't there... | 17:24 |
| *** cody-somerville has joined #openstack-horizon | 17:25 | |
| david-lyle | sambetts: around? | 17:26 |
| david-lyle | lhcheng, yeah, this is related to the issue yesterday | 17:28 |
| sambetts | david-lyle: Hi! | 17:31 |
| thiagop | lhcheng: do you mean the issue of couldn't change between projects? | 17:31 |
| david-lyle | we're discussing a patch of yours | 17:31 |
| thiagop | s/change/changing/ | 17:32 |
| sambetts | The panel access caching one I'm, guessing?? | 17:32 |
| david-lyle | https://github.com/openstack/horizon/commit/16db58fabad8934b8fbdfc6aee0361cc138b20af | 17:32 |
| david-lyle | yeah | 17:32 |
| david-lyle | it's problematic on a few levels | 17:32 |
| david-lyle | 1) it blows up the session cookie size limit | 17:33 |
| lhcheng | thiagop: yes! | 17:33 |
| *** ericpeterson has quit IRC | 17:33 | |
| david-lyle | 2) it seems to not be updating on token rescope changes | 17:33 |
| thiagop | lhcheng: This is what I've found so far https://bugs.launchpad.net/horizon/+bug/1382079 | 17:34 |
| thiagop | sambetts: ^ | 17:34 |
| david-lyle | we're considering either moving to use memoize, or revert completely and rethink | 17:34 |
| david-lyle | would like your opinion | 17:34 |
| *** aberezin has quit IRC | 17:36 | |
| sambetts | I've not used memoize before so I can't comment there, in response to your second point I'm pretty sure I tested that senario and it was updating when I change the project from the drop down so I'm not sure what going on there | 17:37 |
| david-lyle | sambetts, memoize would store the results in memory on the server rather than in the session cookie | 17:38 |
| sambetts | as far as I could see it wasn't storing it in the cookie anyway, I see a session key, but I never saw anything else stored in the cookie, surly I would have seen an error as soon as it hit the 4k limit | 17:40 |
| sambetts | I assumed it used the key in the cookie to access something in memory | 17:41 |
| *** tnovacik has quit IRC | 17:41 | |
| *** ArcTanSusan has quit IRC | 17:45 | |
| *** athomas has quit IRC | 17:46 | |
| thiagop | sambetts david-lyle : according to ericpetterson, the problem doesn't occurs when you are using the SQL session backend | 17:46 |
| sambetts | which is default? because I run standard devstack and have not seen any issues | 17:47 |
| *** arborism has joined #openstack-horizon | 17:47 | |
| thiagop | I believe the default is signed_cookie. | 17:48 |
| *** arborism has quit IRC | 17:48 | |
| thiagop | sambetts: did you run it using keystone v2.0 or v3? | 17:48 |
| sambetts | whichever is default in devstack | 17:48 |
| thiagop | david-lyle: ^ | 17:49 |
| sambetts | the only config I've changed in the devstack env is to turn on neutron | 17:49 |
| sambetts | everything else I've left default | 17:49 |
| thiagop | maybe the session is blowing up due to v3 token info | 17:49 |
| tsufiev | david-lyle, what if we look in the direction of clientside localStorage? I've been thinking about whether could solve the problems that cookie-session have? | 17:50 |
| sambetts | thiagop: In reference to that bug report, that seems to me like something else is falling out of sync if its passing the wrong context into the decorator | 17:52 |
| thiagop | sambetts: I believe it too. But I've looking into it a day and a half and cannot yet find where it's changing this context | 17:53 |
| *** wuhg has quit IRC | 17:53 | |
| thiagop | but, as david-lyle said, it's putting too much info into the session cookie | 17:53 |
| thiagop | david-lyle lhcheng sambetts the problem indeed doesn't occurs when using identity v2.0 | 17:54 |
| *** ArcTanSusan has joined #openstack-horizon | 17:56 | |
| lhcheng | thiagop identity v2 has smaller footprint on the data stored in session than v3 | 17:57 |
| *** cody-somerville has quit IRC | 17:57 | |
| *** clu_ has joined #openstack-horizon | 17:57 | |
| lhcheng | thiagop, the endpoint data is smaller, due to the data structure difference | 17:58 |
| openstackgerrit | Jiri Tomasek proposed a change to openstack/horizon: Restyled sidebar to resemble UX guidelines https://review.openstack.org/126289 | 17:58 |
| sambetts | I confused by the difference between using .session and .cookie then, I thought that .session wasn't bound by the 4k limit | 17:58 |
| lhcheng | as thiagop mentioned, the default is signed_cookies for storing session information | 17:59 |
| lhcheng | the session information is passed back and forth as cookies | 17:59 |
| thiagop | sambetts: the session, by default, is stored in cookies. | 17:59 |
| sambetts | I hadn't realised that it was stored as a encrypted blob, I see that now, I thought that the hash I was seeing was just a unique key back to something stored in memory on the server | 18:00 |
| thiagop | sambetts: that's when you're using database session backend | 18:01 |
| thiagop | btw, why do we use signed_cookies as default? | 18:02 |
| *** ericpeterson has joined #openstack-horizon | 18:04 | |
| sambetts | if we use memoize does that use a key stored in the cookie/session for retreival?? | 18:04 |
| lhcheng | thiagop: I guess it was used to simplify devstack setup, less dependency (no additional service). | 18:04 |
| sambetts | keystone v3 uses the tokens that are close to 4k anyway, so surely that leads to nearly all other cookie/session storage becoming problematic? Like the user settings for instance | 18:06 |
| *** akrivoka has joined #openstack-horizon | 18:07 | |
| *** ericpeterson has quit IRC | 18:08 | |
| openstackgerrit | Nikita Konovalov proposed a change to openstack/horizon: [Data_processing] fix job_executions table https://review.openstack.org/125927 | 18:09 |
| openstackgerrit | Cindy Lu proposed a change to openstack/horizon: add Column attribute for help tooltip in heading https://review.openstack.org/128804 | 18:09 |
| *** akrivoka has quit IRC | 18:14 | |
| *** e0ne has quit IRC | 18:16 | |
| *** jwy has joined #openstack-horizon | 18:16 | |
| thiagop | david-lyle: changing the subject for a while, we have made a PoC of using the endpoint policy to do the RBAC validation on Horizon | 18:18 |
| thiagop | david-lyle: do you mind to take a look? | 18:18 |
| openstackgerrit | Sayali Lunkad proposed a change to openstack/horizon: Removes term inconsistency in the dashboard https://review.openstack.org/73222 | 18:22 |
| *** ericpeterson has joined #openstack-horizon | 18:23 | |
| *** jwy1 has joined #openstack-horizon | 18:25 | |
| *** jwy has quit IRC | 18:25 | |
| david-lyle | thiagop, additionally, HA is much easier when passing all the session info around. | 18:25 |
| david-lyle | what really needs to happen is the service catalog getting removed from the token | 18:26 |
| *** jwy1 has left #openstack-horizon | 18:26 | |
| david-lyle | pki tokens will still be a problem then | 18:26 |
| ericpeterson | if we can assume everyone's catalog always looks the same, then I'm on board | 18:26 |
| david-lyle | think separate API | 18:27 |
| thiagop | david-lyle: It's simple. There is a flag "--nocatalog" on getting tokens on keystone | 18:27 |
| david-lyle | moving forward | 18:28 |
| david-lyle | and the HTTP interface supports that? | 18:28 |
| thiagop | you mean the REST API? | 18:30 |
| david-lyle | via keystoneclient yes | 18:31 |
| thiagop | give me 5min to check... | 18:31 |
| david-lyle | not seeing it documented in the client, that may just be an oversight | 18:32 |
| thiagop | on the client, it seems that it's not available | 18:34 |
| thiagop | but we can take care of it :D | 18:35 |
| david-lyle | I'm sure that will change, but hard to use as is :) | 18:35 |
| *** ramishra has joined #openstack-horizon | 18:35 | |
| *** johnma has joined #openstack-horizon | 18:37 | |
| *** ArcTanSusan has quit IRC | 18:38 | |
| *** ramishra has quit IRC | 18:40 | |
| *** ramishra has joined #openstack-horizon | 18:53 | |
| *** ramishra has quit IRC | 18:58 | |
| *** ArcTanSusan has joined #openstack-horizon | 18:59 | |
| thiagop | sambetts david-lyle lhcheng ericpeterson : what do you think of this structure? http://paste.openstack.org/show/121522/ | 19:01 |
| *** sbfox has joined #openstack-horizon | 19:01 | |
| thiagop | seems to work (left the prints for you to see) | 19:01 |
| thiagop | when you open a window for the first time, it reloads the cache and, if you reload the same page on the same project, it doesn't reloads | 19:02 |
| thiagop | if you change the project, the reloading occurs | 19:02 |
| thiagop | (there is also a CACHE global on top of the module) | 19:02 |
| *** ybathia has quit IRC | 19:06 | |
| *** ybathia has joined #openstack-horizon | 19:09 | |
| *** cody-somerville has joined #openstack-horizon | 19:09 | |
| *** jprovazn_afk has quit IRC | 19:12 | |
| *** ybathia has quit IRC | 19:14 | |
| openstackgerrit | Adrien Vergé proposed a change to openstack/horizon: [Sahara] Support Cinder availability zones https://review.openstack.org/128906 | 19:15 |
| sambetts | thiagop: does every user get a clean CACHE variable? or does the global go between users?? | 19:15 |
| thiagop | sambetts: I didn't test that, but I believe each user has a cache due to threading isolation. | 19:16 |
| thiagop | sambetts: But I have to test that indeed | 19:17 |
| *** ArcTanSusan has quit IRC | 19:18 | |
| sambetts | I don't think every Django communication/session runs in isolation so I think you would end up with cross talk between clients, thats why we have the session store | 19:18 |
| sambetts | otherwise you could just store everything in variables and you would never need the cookie | 19:19 |
| *** david-lyle_ has joined #openstack-horizon | 19:20 | |
| *** david-lyle has quit IRC | 19:20 | |
| thiagop | sambetts: makes sense. | 19:21 |
| *** mrunge has quit IRC | 19:23 | |
| thiagop | sambetts: maybe it works because the token is different between users. But I wonder how users may "steal" the cache from each other in a production environment... | 19:25 |
| *** david-lyle_ is now known as david-lyle | 19:27 | |
| sambetts | I think what will happen right now with your suggestion, is that every time a different user refreshes the page it invalidates the cache and rebuilds it so there is no advantage unless you only have one user | 19:27 |
| sambetts | thiagop: ^ | 19:27 |
| sambetts | thiagop: I can see that leading into some serious race hazards too, its not somewhere we want to go | 19:28 |
| *** exploreshaifali has joined #openstack-horizon | 19:28 | |
| thiagop | sambetts: so, we have to find another way of approaching this without putting the cached checks into the session. | 19:29 |
| openstackgerrit | David Lyle proposed a change to openstack/horizon: Fixing template syntax https://review.openstack.org/129028 | 19:29 |
| sambetts | thiagop: agreed, something that I was considering was, similar to yours except the CACHE variable had another layer above it so it was like { some_session_identifier : the_cached_data, another_session_identifier: thier cached data } however what I've realised is that using variables/global or is just a plain old bad idea for scalablity of the dashboard, | 19:34 |
| *** clu_ has quit IRC | 19:35 | |
| sambetts | say your runing multiple horizons under a load balancer variables won't be shared == bad idea | 19:35 |
| *** ArcTanSusan has joined #openstack-horizon | 19:37 | |
| sambetts | I've have been thinking about the cookie idea for a bit now, because of another patch I worked on, and I have learned that you can have multple cookies for a single site, which woul allow us to extend past the 4k limit, then we could have a cookie for the token/auth stuff, and one for user settings and one for something like this? | 19:39 |
| sambetts | if I've understood it correctly that is, | 19:39 |
| *** absubram has quit IRC | 19:39 | |
| sambetts | thiagop, david-lyle: ^ | 19:39 |
| *** ArcTanSusan has quit IRC | 19:41 | |
| david-lyle | sambetts, regardless of the number, size limit is 4K | 19:41 |
| david-lyle | even if you have 50 cookies, they share the same 4K cap | 19:42 |
| ericpeterson | must be site wide ? | 19:42 |
| sambetts | even with mutliple cookie files, so they are bound to different urls? | 19:42 |
| david-lyle | yes | 19:42 |
| david-lyle | per domain | 19:42 |
| david-lyle | I believe | 19:42 |
| david-lyle | looked into the details many times | 19:43 |
| ericpeterson | I'd say the caching right now is not 100%, and for many uses you are best off not caching at all | 19:43 |
| ericpeterson | it's better to have a slower system and you can actually switch projects, than to have a fast setup with only one project | 19:44 |
| sambetts | the cache really isn't the issue tbh, the issue is that once you introduce the PKI token you have zero storage left for anything let alone caching | 19:45 |
| ericpeterson | well, the cache places 43 or something like that items into the cookie, so I'd say it's part of the problem | 19:46 |
| *** cody-somerville has quit IRC | 19:46 | |
| ericpeterson | and if you remove the cache, everything else works.... so I'd disagree and say the cache most certainly is part of the problem | 19:46 |
| ericpeterson | but yeah, it's resource contention in the big picture | 19:46 |
| *** ArcTanSusan has joined #openstack-horizon | 19:47 | |
| sambetts | admittly there are improvments that need to be made and can be made in the future, but once v3 comes in I think it even proves an issue for storing the user settings | 19:47 |
| lhcheng | sambetts, we got a workaround for the PKI token. It is a hack, but it works with horizon in devstack. | 19:48 |
| sambetts | oh cool! have you got a link? | 19:49 |
| david-lyle | we should not be pushing a bunch of method references in the session | 19:51 |
| sambetts | david-lyle: When I was working on it, I discussed a lot here, and we decided that right now it was the only way to reference each panel uniquly as there is no unqiue identifier for each panel | 19:53 |
| david-lyle | how does that translate if your session gets load balanced to another server/ | 19:53 |
| david-lyle | ? | 19:53 |
| *** ramishra has joined #openstack-horizon | 19:54 | |
| sambetts | they are class refrences as far as im aware the class names and locations wont change dynamically | 19:55 |
| sambetts | or based on different instances | 19:55 |
| sambetts | they aren't object references | 19:55 |
| david-lyle | I haven't looked at how it's encoded | 19:56 |
| sambetts | "openstack_dashboard.dashboard.project.network" will always be just that afaik | 19:56 |
| *** romainh has quit IRC | 19:56 | |
| david-lyle | ok | 19:56 |
| *** e0ne has joined #openstack-horizon | 19:57 | |
| sambetts | I would imagen if you wanted to reduce the size that it takes up in the cookie, there is alot of redundency in the names so it would compress well | 19:58 |
| *** cody-somerville has joined #openstack-horizon | 19:58 | |
| *** ramishra has quit IRC | 19:59 | |
| sambetts | or something I discuss with Radomir I think was actually adding a variable to every horizon_component object which is a unique identifier, like "admin_network_panel" or something similar | 19:59 |
| sambetts | and have a check similar to the slug check that ensures that one is given to each component and that its unique with every other one that exists so far | 20:00 |
| sambetts | then you wouldn't need the full class path for everything | 20:01 |
| *** clu_ has joined #openstack-horizon | 20:02 | |
| sambetts | is there a reason that the slug variable isn't unique? Every panel has to have one but for instance both admin->network and project->network both have the slug "networks" | 20:03 |
| *** david-lyle_ has joined #openstack-horizon | 20:04 | |
| *** e0ne has quit IRC | 20:05 | |
| *** david-lyle has quit IRC | 20:06 | |
| sambetts | david-lyle_: ^ | 20:07 |
| ericpeterson | so keeping your permissions in the clients cookie is an idea that seems to be pretty sticky here | 20:09 |
| sambetts | ericpeterson: its only the same as the token is in the cookie because its in the session in this case | 20:11 |
| openstackgerrit | Sam Betts proposed a change to openstack/horizon: Pass correct project ID to get tenant_usages https://review.openstack.org/129037 | 20:13 |
| ericpeterson | we had a problem that sometimes happened, where we would overflow the cookie size from time to time. now it's worse / more frequent. that much we all agree upon, right? | 20:15 |
| *** ybathia has joined #openstack-horizon | 20:17 | |
| sambetts | I think cookie based session store is a bad idea period, for security and for the fact that people understand that cookies have a limit and sessions normally dont, and storing the session in the cookie just limits our use of the session, surly if we want a string free implmentation for devstack we can just use Sqlite or something as the session backend | 20:18 |
| thiagop | ericpeterson: +1 | 20:18 |
| thiagop | sambetts: I also think that cookie transmission is a undesirable overhead on the traffic user<->web server | 20:20 |
| ericpeterson | sambetts: I agree that cookie backed sessions are not realistic, but that's jumping ahead I think. I'd like to move to something else, but that's not a 2 day fix either | 20:20 |
| thiagop | It's not a "oh my god, this cookie is huge", but if you have a clean solution that could do the same thing without transmitting this "almost 4K cookie", it would be awesome | 20:21 |
| sambetts | ericpeterson: everything uses the django helpers for session and cookie so switching the default backend to something else shouldnt effect the code at all, something like Sqlite should just drop in no hassel | 20:22 |
| *** ybathia has quit IRC | 20:22 | |
| sambetts | django will handle the rest surly | 20:22 |
| sambetts | even just swapping it out to django.contrib.sessions.backends.file would be more desirable wouldn't it | 20:23 |
| *** woodm1979 has joined #openstack-horizon | 20:24 | |
| *** sambetts_ has joined #openstack-horizon | 20:25 | |
| *** sambetts has quit IRC | 20:26 | |
| thiagop | we have the 'file', 'cache' and 'cached_db' options. I tried with cached_db and I throws an exception because settings.DATABASES is not configured | 20:26 |
| david-lyle_ | thiagop, other session types require additional configuration | 20:27 |
| thiagop | I know that sqlite has a way to store the database in [memory] instead of in a file, but I'm worried about the concurrent access... | 20:27 |
| thiagop | david-lyle_: 'cache' doesn't | 20:28 |
| sambetts_ | sqlite knows how to handle concurrency afaik | 20:28 |
| david-lyle_ | sqlite used to be the default | 20:28 |
| david-lyle_ | in Horizon back in Grizzly I believe | 20:29 |
| david-lyle_ | requires an additional call to manage.py to syncdb on set up | 20:30 |
| sambetts_ | back then keystone used PKI too didn't it? I remeber working with it a while ago and PKI was causing issues with cookies for me | 20:30 |
| david-lyle_ | sambetts_, maybe, can't remember when PKI support was added, Grizzly or Havana | 20:31 |
| *** ArcTanSusan has quit IRC | 20:31 | |
| sambetts_ | I ended up have to move everything into SQLite in my project to handle the amount of data I wanted to store alongside the token | 20:32 |
| sambetts_ | was there a reason that horizon moved away from SQLite ?? | 20:32 |
| openstackgerrit | Sayali Lunkad proposed a change to openstack/horizon: Removes term inconsistency in the dashboard https://review.openstack.org/73222 | 20:34 |
| ericpeterson | I think it's that one additional call on manage syncdb | 20:36 |
| ericpeterson | i.e. convenience | 20:36 |
| ericpeterson | not such a good idea, as we can see now ;) | 20:36 |
| thiagop | ericpeterson: in fact, it's a sine qua non condition to have a database on horizon. | 20:38 |
| thiagop | so far, we rely the persistence of everything in other services.. | 20:38 |
| sambetts_ | ericpeterson: thats something that could be easily dealt with in the devstack script or run_tests right? | 20:39 |
| ericpeterson | yes, it could | 20:39 |
| ericpeterson | it also raises a different issue around HA | 20:40 |
| ericpeterson | if you used sqllite in the file mode, that is not really going to work for HA | 20:40 |
| ericpeterson | if you use cookies, that *can* work for HA (so long as you have one region, keep your cookie size small, etc | 20:41 |
| ericpeterson | so some of the openstack distros / vendors might care about this | 20:41 |
| sambetts_ | I think at the point you start running a HA environment youll want to be running a real DB, every other service is so why does horizon need to be different | 20:42 |
| ericpeterson | agreed | 20:44 |
| *** cody-somerville has quit IRC | 20:46 | |
| *** david-lyle_ is now known as david-lyle | 20:48 | |
| thiagop | sambetts_: +1 | 20:49 |
| *** sambetts has joined #openstack-horizon | 20:49 | |
| *** sambetts is now known as Guest7418 | 20:49 | |
| *** crobertsrh is now known as _crobertsrh | 20:50 | |
| *** sambetts_ has quit IRC | 20:51 | |
| *** Guest7418 is now known as sambetts_ | 20:51 | |
| *** amcrn has joined #openstack-horizon | 20:53 | |
| sambetts_ | ericpeterson, david-lyle: My internet died for some reason, while I was mid reply, I'm heading off for the evening, its nearly 22:00 where I am, I'll be back online ~9:00 UTC so can we continue these discussions tomrrow? | 20:53 |
| david-lyle | sure | 20:54 |
| david-lyle | have a good night | 20:54 |
| thiagop | sambetts_: sure, good night! | 20:54 |
| sambetts_ | Thanks v.much see you tomorrow | 20:54 |
| *** ramishra has joined #openstack-horizon | 20:55 | |
| *** sambetts_ has quit IRC | 20:55 | |
| thiagop | david-lyle: have you taken a look in the PoC I mentioned before? | 20:58 |
| *** raildo is now known as raildo_away | 20:59 | |
| david-lyle | thiagop, not yet, it's in my queue | 20:59 |
| *** ramishra has quit IRC | 20:59 | |
| *** cbader has joined #openstack-horizon | 21:00 | |
| thiagop | Ok. Just to know that you have seen and is not forgotten. No pressure. :) | 21:00 |
| ericpeterson | thanks, good night :) | 21:00 |
| *** ArcTanSusan has joined #openstack-horizon | 21:01 | |
| thiagop | ericpeterson: thanks to you! see ya. | 21:02 |
| *** ArcTanSusan has quit IRC | 21:02 | |
| *** doug-fish has quit IRC | 21:03 | |
| *** ArcTanSusan has joined #openstack-horizon | 21:07 | |
| *** thiagop has quit IRC | 21:08 | |
| *** ybathia has joined #openstack-horizon | 21:13 | |
| *** exploreshaifali has quit IRC | 21:23 | |
| *** cody-somerville has joined #openstack-horizon | 21:25 | |
| *** cody-somerville has quit IRC | 21:25 | |
| *** cody-somerville has joined #openstack-horizon | 21:25 | |
| *** nlahouti has joined #openstack-horizon | 21:26 | |
| *** sayali has quit IRC | 21:31 | |
| openstackgerrit | David Lyle proposed a change to openstack/horizon: Juno Release Notes https://review.openstack.org/128744 | 21:33 |
| *** cody-somerville has quit IRC | 21:35 | |
| *** MaxV has joined #openstack-horizon | 21:36 | |
| *** freyes has quit IRC | 21:40 | |
| *** freyes has joined #openstack-horizon | 21:41 | |
| *** MaxV has quit IRC | 21:41 | |
| *** alexpilotti has joined #openstack-horizon | 21:41 | |
| *** sayali has joined #openstack-horizon | 21:44 | |
| *** MaxV has joined #openstack-horizon | 21:45 | |
| *** julim has quit IRC | 21:47 | |
| *** MaxV_ has joined #openstack-horizon | 21:48 | |
| *** david-lyle has quit IRC | 21:49 | |
| *** MaxV has quit IRC | 21:49 | |
| *** ramishra has joined #openstack-horizon | 21:55 | |
| *** MaxV has joined #openstack-horizon | 21:59 | |
| *** cody-somerville has joined #openstack-horizon | 22:00 | |
| *** cody-somerville has quit IRC | 22:00 | |
| *** cody-somerville has joined #openstack-horizon | 22:00 | |
| *** ramishra has quit IRC | 22:00 | |
| *** MaxV_ has quit IRC | 22:02 | |
| *** radez is now known as radez_g0n3 | 22:04 | |
| *** ArcTanSusan has quit IRC | 22:04 | |
| *** david-lyle has joined #openstack-horizon | 22:05 | |
| *** johnma has quit IRC | 22:06 | |
| *** rwsu has quit IRC | 22:07 | |
| *** ArcTanSusan has joined #openstack-horizon | 22:07 | |
| *** packet has quit IRC | 22:09 | |
| *** ericpeterson has quit IRC | 22:20 | |
| *** TravT has quit IRC | 22:23 | |
| *** david-lyle has quit IRC | 22:24 | |
| *** woodm1979 has quit IRC | 22:24 | |
| *** afaranha has quit IRC | 22:24 | |
| rodrigods | amotoki, lhcheng there? | 22:25 |
| *** peristeri has quit IRC | 22:27 | |
| *** colinmcnamara has joined #openstack-horizon | 22:28 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 22:28 | |
| *** MaxV has quit IRC | 22:28 | |
| *** afaranha has joined #openstack-horizon | 22:30 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 22:37 | |
| lhcheng | rodrigods, yes? | 22:38 |
| *** colinmcnamara has quit IRC | 22:39 | |
| rodrigods | lhcheng, not sure how the backport potential works, but I think that maybe is the case with: https://bugs.launchpad.net/horizon/+bug/1369137 | 22:39 |
| lhcheng | not sure if it is really backport potential, don't think people have really started using keystone v3 :P | 22:40 |
| rodrigods | lhcheng, hehe, ok... since federation and k2k are the main "stars" from keystone in juno release, I thought would be the case here | 22:41 |
| lhcheng | even with federation, users are not required to move to v3 (I think) | 22:43 |
| rodrigods | lhcheng, federation is an v3 extension | 22:46 |
| rodrigods | (it requires domains and groups) | 22:46 |
| lhcheng | ah.. another bad news. horizon doesn't support federation yet. | 22:47 |
| rodrigods | yeah, hope it will soon | 22:47 |
| lhcheng | hopefully in kilo | 22:47 |
| lhcheng | yeah | 22:47 |
| *** ramishra has joined #openstack-horizon | 22:56 | |
| *** ramishra has quit IRC | 23:01 | |
| *** sayali has quit IRC | 23:03 | |
| *** cbader has quit IRC | 23:15 | |
| *** sbfox has quit IRC | 23:24 | |
| *** echevemaster has joined #openstack-horizon | 23:24 | |
| *** cody-somerville has quit IRC | 23:39 | |
| *** exploreshaifali has joined #openstack-horizon | 23:40 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 23:52 | |
| *** ramishra has joined #openstack-horizon | 23:57 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!