| mgagne | so far... in horizon.decorators request.user=AnonymousUser when calling require_auth. that's all I found for now. | 00:04 |
|---|---|---|
| adriant | mgagne: At the very least, open a bug for it with the notes of what you've found | 00:32 |
| mgagne | I think I found the issue, just tracking the git history | 00:32 |
| adriant | ooo?! | 00:32 |
| mgagne | https://github.com/openstack/horizon/blob/master/openstack_auth/backend.py#L72 | 00:33 |
| mgagne | request is not available and there is a note about monkey patching | 00:33 |
| *** betherly has joined #openstack-horizon | 00:35 | |
| mgagne | usage of self.request goes up to original commit in django_openstack_auth: https://github.com/openstack/django_openstack_auth/commit/9962375e0c5209ff82b7d6b5fda92bfec8dc4e1c | 00:35 |
| mgagne | so I'm not sure how it ever worked | 00:36 |
| mgagne | because once you are logged, it looks like you don't go through that code path, can't find my debug log | 00:36 |
| mgagne | there is patch_middleware_get_user in openstack_auth which monkey patches the get_user of the backend and injects the request attribute. I'm not sure what's supposed to call it. | 00:39 |
| adriant | interesting | 00:39 |
| mgagne | but your name is in the comments so I feel you should know something :P | 00:39 |
| *** betherly has quit IRC | 00:39 | |
| mgagne | https://github.com/openstack/horizon/blob/master/openstack_auth/utils.py#L63 | 00:39 |
| mgagne | called here https://github.com/openstack/horizon/blob/master/openstack_auth/urls.py#L20 | 00:40 |
| adriant | huh | 00:41 |
| adriant | so it is | 00:41 |
| adriant | Oh, I know why. It's because I was doing the work to migrate between Django versions | 00:41 |
| mgagne | and here https://github.com/openstack/horizon/blob/master/openstack_dashboard/settings.py#L196 | 00:41 |
| mgagne | I don't know what in Django/Horizon loads it | 00:42 |
| adriant | https://github.com/openstack/django_openstack_auth/commit/601e1ad44c18b9665c2ec6886438299de72d3761 | 00:43 |
| mgagne | well, I do use Django 1.11 and it's broken ¯\_(ツ)_/¯ | 00:43 |
| adriant | I'm curious if this bug existed in earlier versions | 00:44 |
| mgagne | https://github.com/openstack/horizon/blob/master/openstack_dashboard/urls.py#L52 | 00:45 |
| adriant | so it does get called | 00:46 |
| adriant | just, in a very roundabout way | 00:46 |
| mgagne | nope, only "introspected" | 00:46 |
| adriant | OH | 00:46 |
| adriant | ffs | 00:47 |
| mgagne | but if you access any pages under auth/ in a different browser, you don't get the issue anymore | 00:47 |
| adriant | so until it gets called at least once, that code path isn't executed | 00:47 |
| mgagne | because now it gets loaded for real | 00:47 |
| mgagne | yea | 00:47 |
| adriant | so we need to move that function call somewhere else | 00:47 |
| mgagne | I suppose ;) | 00:47 |
| *** cshen has joined #openstack-horizon | 00:50 | |
| *** cshen has quit IRC | 00:54 | |
| *** betherly has joined #openstack-horizon | 00:56 | |
| adriant | nope | 00:57 |
| adriant | tried moving it to openstack_dashboard/urls.py | 00:58 |
| adriant | same error | 00:58 |
| mgagne | I tried to move it around and I get other errors from Django | 01:01 |
| *** betherly has quit IRC | 01:01 | |
| mgagne | could be that django.contrib.auth.middleware.AuthenticationMiddleware is already loaded at that moment | 01:03 |
| mgagne | but at this point, I'm like that dog in the lab, I don't know what I'm doing. | 01:04 |
| adriant | yeah, that's about where I got to when chasing this last time. :/ | 01:06 |
| mgagne | so django.middleware.common.CommonMiddleware is early enough but might not be the right place. :P | 01:06 |
| mgagne | I might just create a dummy middleware and load it from there | 01:08 |
| mgagne | someone else had a similar idea https://opensourcehacker.com/2010/08/12/applying-monkey-patches-in-django-middleware-layer/ | 01:09 |
| mgagne | I'm gonna be that guy https://gist.github.com/mgagne/185776102e73766571978ac295861baa | 01:13 |
| *** dayou has quit IRC | 01:38 | |
| adriant | does it work? | 01:47 |
| mgagne | yes | 01:47 |
| mgagne | I can restart with 20 processes, no issue. | 01:48 |
| *** betherly has joined #openstack-horizon | 01:48 | |
| mgagne | just going to reset my environment to make sure | 01:48 |
| mgagne | all looks fine, currently repackaging on my side | 01:50 |
| adriant | just adding the same to my code and will test | 01:51 |
| adriant | where in the middleware classes list did you put it? | 01:51 |
| mgagne | 1st one | 01:51 |
| *** betherly has quit IRC | 01:52 | |
| adriant | works! | 01:55 |
| adriant | I've limited gunicorn to 1 worker, and am having it restart after 5 requests | 01:55 |
| adriant | i've not seen a logout | 01:56 |
| mgagne | =) | 01:56 |
| adriant | mgagne: I expect to see a patch asap, and I would still file a bug report to help the horizon team understand the bug | 01:58 |
| adriant | mgagne: thanks sooooo much | 01:59 |
| adriant | I can't wait to port this to our build | 01:59 |
| mgagne | hehe, it bothers me too | 01:59 |
| adriant | mostly because... horizon has a memory leak problem, and we solve that by restarting it often... | 02:00 |
| adriant | which makes our users get "logged out" | 02:00 |
| mgagne | ouch | 02:00 |
| mgagne | yea, we had to restart Horizon too, OOM. but running Kilo in some environments. | 02:00 |
| adriant | so the memory problem is mostly due to the memoized cached | 02:01 |
| mgagne | there is a bug already: https://bugs.launchpad.net/horizon/+bug/1764622 | 02:01 |
| openstack | Launchpad bug 1764622 in OpenStack Dashboard (Horizon) "Restarting the web server causes users to get kicked out" [Undecided,Confirmed] | 02:01 |
| adriant | and the solution (sort of) to the memory leak issue: https://review.openstack.org/#/c/605024/ and https://review.openstack.org/#/c/608801 | 02:02 |
| adriant | oh cool, not sure how I missed that bug | 02:02 |
| openstackgerrit | Adrian Turjak proposed openstack/horizon master: rework memoized as a LRU cache with expiry https://review.openstack.org/608801 | 02:12 |
| mgagne | ok, added details in bug, I will propose a fix tomorrow as it's 10pm here. you can propose a fix too if you want or have a better idea. | 02:17 |
| adriant | nah, that fix is perfect | 02:19 |
| adriant | mgagne: but do add me to review :) | 02:19 |
| adriant | and if the memory leak affects you, feel free to review/test the above two patches | 02:20 |
| mgagne | it took a couple of weeks/months to get the OOM :P | 02:20 |
| mgagne | or maybe it crashed and restarted by itself with Puppet. who knows | 02:21 |
| adriant | ah, and no many users probably? | 02:22 |
| adriant | not* | 02:22 |
| openstackgerrit | Mathieu Gagné proposed openstack/horizon master: Fix django.contrib.auth.middleware monkey patching https://review.openstack.org/614673 | 02:36 |
| mgagne | we have multiple web servers too so I don't know for sure. | 02:37 |
| openstackgerrit | Mathieu Gagné proposed openstack/horizon master: Fix django.contrib.auth.middleware monkey patching https://review.openstack.org/614673 | 02:43 |
| *** Leo_m has quit IRC | 02:45 | |
| *** Leo_m has joined #openstack-horizon | 02:50 | |
| *** cshen has joined #openstack-horizon | 02:51 | |
| *** cshen has quit IRC | 02:55 | |
| *** betherly has joined #openstack-horizon | 03:21 | |
| *** betherly has quit IRC | 03:26 | |
| *** btully has joined #openstack-horizon | 03:38 | |
| *** btully has quit IRC | 03:42 | |
| *** betherly has joined #openstack-horizon | 03:53 | |
| openstackgerrit | Andy Botting proposed openstack/horizon master: Support setting more visibility options for edit/create image https://review.openstack.org/602468 | 03:55 |
| openstackgerrit | Andy Botting proposed openstack/horizon master: Better support for community images https://review.openstack.org/614688 | 03:55 |
| *** betherly has quit IRC | 03:57 | |
| openstackgerrit | Andy Botting proposed openstack/horizon master: Support setting more visibility options for edit/create image https://review.openstack.org/602468 | 03:58 |
| *** dayou has joined #openstack-horizon | 03:59 | |
| *** betherly has joined #openstack-horizon | 04:24 | |
| *** betherly has quit IRC | 04:29 | |
| *** cshen has joined #openstack-horizon | 04:51 | |
| *** cshen has quit IRC | 04:56 | |
| openstackgerrit | Adrian Turjak proposed openstack/horizon master: Fix django.contrib.auth.middleware monkey patching https://review.openstack.org/614673 | 04:58 |
| adriant | mgagne: I've submitted some additional cleanup to your patch rather than leaving notes because I thought it would make your life a little easier than chasing comments. :) | 05:12 |
| *** betherly has joined #openstack-horizon | 05:16 | |
| *** betherly has quit IRC | 05:21 | |
| *** ircuser-1 has quit IRC | 05:24 | |
| *** btully has joined #openstack-horizon | 05:26 | |
| *** btully has quit IRC | 05:31 | |
| openstackgerrit | Adrian Turjak proposed openstack/horizon master: IGNORE THIS https://review.openstack.org/614697 | 05:37 |
| *** betherly has joined #openstack-horizon | 05:37 | |
| *** betherly has quit IRC | 05:42 | |
| *** betherly has joined #openstack-horizon | 05:57 | |
| *** betherly has quit IRC | 06:02 | |
| *** cshen has joined #openstack-horizon | 06:15 | |
| *** cshen has quit IRC | 06:21 | |
| *** cshen has joined #openstack-horizon | 06:43 | |
| *** cshen has quit IRC | 07:08 | |
| *** cshen has joined #openstack-horizon | 07:14 | |
| *** cshen has quit IRC | 07:19 | |
| *** cshen has joined #openstack-horizon | 07:34 | |
| *** cshen has quit IRC | 07:39 | |
| *** pcaruana|elisa| has joined #openstack-horizon | 07:40 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/horizon master: Imported Translations from Zanata https://review.openstack.org/614704 | 07:50 |
| *** pcaruana|elisa| has quit IRC | 07:59 | |
| *** pcaruana has joined #openstack-horizon | 08:05 | |
| *** cshen has joined #openstack-horizon | 08:06 | |
| *** cshen has quit IRC | 08:38 | |
| *** cshen has joined #openstack-horizon | 08:48 | |
| *** rdopiera has joined #openstack-horizon | 08:55 | |
| *** jpich has joined #openstack-horizon | 09:00 | |
| *** e0ne has joined #openstack-horizon | 09:43 | |
| *** ChanServ sets mode: +o e0ne | 09:43 | |
| *** v12aml has joined #openstack-horizon | 09:44 | |
| *** phuongnh has joined #openstack-horizon | 10:15 | |
| *** phuongnh has quit IRC | 10:31 | |
| *** e0ne has quit IRC | 10:32 | |
| *** e0ne has joined #openstack-horizon | 10:36 | |
| *** ChanServ sets mode: +o e0ne | 10:36 | |
| *** jtomasek has quit IRC | 10:43 | |
| *** jtomasek has joined #openstack-horizon | 10:45 | |
| *** dave-mccowan has joined #openstack-horizon | 10:45 | |
| *** btully has joined #openstack-horizon | 10:53 | |
| *** btully has quit IRC | 10:57 | |
| *** pcaruana has quit IRC | 11:05 | |
| *** cshen has quit IRC | 11:50 | |
| *** cshen has joined #openstack-horizon | 11:52 | |
| *** pcaruana has joined #openstack-horizon | 11:52 | |
| mgagne | adriant: go ahead | 12:25 |
| adriant | mgagne: I already did, and it seems to have actually fixed a failure :) | 12:26 |
| adriant | to test that I pushed https://review.openstack.org/#/c/614697/ to see if the same failure happened again, and it did | 12:27 |
| mgagne | was it mostly related to how tests were done? | 12:27 |
| adriant | more so to how the middleware was used | 12:27 |
| adriant | it needed to be able to be called | 12:28 |
| mgagne | in tests right? because I didn't see any issue so far in my setup | 12:28 |
| adriant | when I included the monkey patching middleware, the tests kept failing because something kept trying to call the middleware | 12:28 |
| adriant | so adding that extra logic to the middleware which does pretty much nothing fixed it | 12:29 |
| mgagne | +1 | 12:29 |
| * adriant isn't exactly sure what he fixed, but it works so *shrug* | 12:29 | |
| adriant | and now we aren't being silly and also calling the middleware in weird places | 12:30 |
| adriant | I mean | 12:30 |
| adriant | calling the monkeypatch | 12:30 |
| mgagne | I'm not sure I fully understand the fix because there is no __init__ in the other middlewares | 12:31 |
| mgagne | hmm | 12:32 |
| mgagne | found one in MiddlewareMixin | 12:32 |
| mgagne | ok, so be it | 12:32 |
| adriant | https://docs.djangoproject.com/en/1.11/topics/http/middleware/ | 12:33 |
| adriant | django's example middleware has it | 12:33 |
| adriant | that's pretty much what I based it on | 12:33 |
| adriant | let me revert my code a sec and rerun the tests, I can show you the failure | 12:33 |
| mgagne | it's fine, I saw the failure yesterday | 12:35 |
| mgagne | something like object() takes no parameters | 12:36 |
| adriant | mgagne: http://paste.openstack.org/show/733773/ | 12:36 |
| adriant | yeah, that | 12:36 |
| mgagne | still weird because I didn't get any issue in unit tests on my side with queens. Django==1.11.9 | 12:38 |
| mgagne | or it doesn't test what I think it's testing :D | 12:38 |
| adriant | it only happened for me when I added the middleware to the openstack_auth test settings.py | 12:40 |
| adriant | but the extra code in the middleware effectively does nothing, so it doesn't matter anyway and I guess is now compatible with whatever different ways it can be used | 12:41 |
| adriant | I'm still just so happy and thankful you found the bug! | 12:42 |
| mgagne | =) finally | 12:42 |
| adriant | with that gone our memory leak problems, while still not idle, are mostly fixed with having uwsgi or gunicorn restart the workers more often. | 12:42 |
| mgagne | now... how hard would it be to fix that "bug" ? ;) https://bugs.launchpad.net/horizon/+bug/1609777 | 12:45 |
| openstack | Launchpad bug 1609777 in OpenStack Dashboard (Horizon) "'can_set_password' ignored in the new launch instance experience" [Medium,In progress] - Assigned to Pascal Boutin (pboutin) | 12:45 |
| adriant | mgagne: amusingly enough... I was actually talking about that recently with my boss. It causes us a bit of pain for windows vms. | 12:47 |
| mgagne | us too, we had to revert to the legacy launch instance | 12:48 |
| adriant | we just get people to setup their windows machines via the web console | 12:48 |
| mgagne | Pascal Boutin is no longer working with us so there won't be an update on that change | 12:49 |
| mgagne | adriant: not really an option because there is a random admin password generated by Nova that Horizon doesn't show on boot. | 12:50 |
| adriant | "As a user, when I launch a new instance, I want to be able to define a password for its default user, so that I can log in to it from the web console." | 12:51 |
| adriant | just digging through our UX related backlog | 12:51 |
| adriant | we actually would be interested in redoing the entirety of the launch instance workflow... but that's not an easy task and it's too low priority for us | 12:52 |
| adriant | huh, this is an interesting one too: | 12:53 |
| mgagne | at this point, I just want an admin password field :P | 12:53 |
| adriant | As a cloud user, when I request the cloud to generate an SSH key on my behalf, I want to download the private key by default and have an option to copy it, so that on my Windows desktop I don't have a broken key. | 12:53 |
| adriant | Acceptance criteria: | 12:53 |
| adriant | Download option must be the default | 12:53 |
| adriant | The plain text key should be hidden by default | 12:53 |
| adriant | If the user clicks on the copy to clipboard action, then the plain text key is shown and the key is copied to the clipboard | 12:53 |
| adriant | we kept having issues with the copy/paste action in horizon for generated keys broke when using tools like putty and such on windows | 12:54 |
| mgagne | what's a broken key? | 12:54 |
| mgagne | oh, right. the key is not in the putty format? | 12:54 |
| adriant | extra spaces, tabs, etc, windows was interpreting it weirdly | 12:54 |
| adriant | the content was fine if you cleaned it up manually, but something about windows and copy/paste doesn't work | 12:55 |
| mgagne | :-/ | 12:55 |
| *** e0ne has quit IRC | 12:55 | |
| mgagne | I think our user base still use/prefer passwords although I agree it might not be for the best. | 12:56 |
| adriant | we mostly try to disallow it, but... windows VMs that's not an option | 12:57 |
| mgagne | so yea, if we could get an admin password field, that would be great ™ | 12:57 |
| adriant | I'll add a note to our ticket for that so if we do anything related to it, to follow up with you as well. | 12:58 |
| mgagne | thanks! | 12:58 |
| adriant | I'm not a horizon core dev though, I just happen to be the person from our company who tends to deal with its code the most | 12:59 |
| mgagne | we have that feature in our backlog as well, marked as impeded for ~1 year =) | 12:59 |
| adriant | for us it's most just soooooo low a priority | 12:59 |
| *** cshen has quit IRC | 13:01 | |
| mgagne | hehe, just found that that you are from catalyst and I think we still have a story in our backlog related to some of your optimizations we need to test. | 13:02 |
| adriant | mgagne: to be fair, a lot of those aren't entirely all that helpful. The problem ultimately is just too many API calls. But at least some I'm vaguely tempted to do a little about if I can get https://review.openstack.org/#/c/608801/ merged. Mostly add a little bit of extra caching globally (extension lists), or per project. | 13:09 |
| adriant | like for example: https://review.openstack.org/#/c/511095/ nova extensions don't change often, we can cache that once, refresh it every 24hrs, and that's maybe one less API call. | 13:10 |
| mgagne | yes, this one was in our list ^ | 13:11 |
| mgagne | but like you, Horizon optimization is very low on our list, mostly best effort. | 13:11 |
| adriant | yep | 13:12 |
| adriant | although in recent weeks... the memory leak has become a priority | 13:12 |
| mgagne | until I'm annoyed enough that I just take the time to fix it | 13:12 |
| adriant | we... have been having to restart horizon 1-2 times a day, and carefully because of that logout bug | 13:13 |
| adriant | we used to just let uwsgi handle it, but we started getting too many users so it restarted workers too often and oddly logged out people | 13:13 |
| mgagne | :-/ | 13:14 |
| adriant | Horizon shouldn't really eat 4gb of memory per host when it's only running 4 uwsgi workers | 13:16 |
| mgagne | ¯\_(ツ)_/¯ | 13:16 |
| adriant | exactly | 13:16 |
| *** belmoreira has joined #openstack-horizon | 13:17 | |
| adriant | anyway, off to sleep. Can you test that my changes still solve the logout issue for you? | 13:19 |
| adriant | and I'll do the same tomorrow with my gunicorn setup | 13:20 |
| mgagne | I can test and report in the change. | 13:22 |
| adriant | mgagne: ty! I'm looking forward to seeing it merged, and backported to our current horizon package | 13:22 |
| mgagne | +1 | 13:22 |
| adriant | well, cherrypicked to our internal repo | 13:22 |
| *** cshen has joined #openstack-horizon | 13:24 | |
| mgagne | small nit for django 1.11, commented in change | 13:29 |
| *** cshen has quit IRC | 13:29 | |
| *** tolisbar has joined #openstack-horizon | 13:34 | |
| *** lblanchard has joined #openstack-horizon | 13:36 | |
| *** cshen has joined #openstack-horizon | 13:40 | |
| *** e0ne has joined #openstack-horizon | 13:40 | |
| *** ChanServ sets mode: +o e0ne | 13:40 | |
| tolisbar | hello all. I've deployed openstack rocky through ansible-kolla 7.0.0 with internal/external networking, and the network_interface, neutron_external_interface, and kolla_external_vip_interface defined in the invenrory. My globals is this(https://pastebin.com/HicQXYaC). When I try to access horizon I get redirected to /horizon/auth/login/?next=/ which is a 404. My horizon.lon and horizon-access.log | 13:44 |
| tolisbar | are here(https://pastebin.com/u9E8hLF0) and here(https://pastebin.com/AN4rM9N6). Any idea what might be causing this ? | 13:44 |
| *** Leo_m has quit IRC | 13:44 | |
| *** cshen has quit IRC | 13:48 | |
| *** e0ne_ has joined #openstack-horizon | 14:09 | |
| *** ChanServ sets mode: +o e0ne_ | 14:09 | |
| *** e0ne has quit IRC | 14:10 | |
| *** Leo_m has joined #openstack-horizon | 14:31 | |
| *** btully has joined #openstack-horizon | 14:31 | |
| *** Leo_m has quit IRC | 14:31 | |
| *** Leo_m has joined #openstack-horizon | 14:32 | |
| *** cshen has joined #openstack-horizon | 14:33 | |
| *** btully has quit IRC | 14:36 | |
| *** mvkr has quit IRC | 14:40 | |
| *** mvkr has joined #openstack-horizon | 15:11 | |
| *** kukacz has quit IRC | 15:12 | |
| *** itlinux has quit IRC | 15:13 | |
| *** kukacz has joined #openstack-horizon | 15:19 | |
| *** gyee has joined #openstack-horizon | 15:33 | |
| *** e0ne_ has quit IRC | 15:34 | |
| *** e0ne has joined #openstack-horizon | 15:34 | |
| *** ChanServ sets mode: +o e0ne | 15:34 | |
| *** belmoreira has quit IRC | 16:11 | |
| *** jpich has quit IRC | 16:13 | |
| *** dnguyen has joined #openstack-horizon | 16:13 | |
| *** itlinux has joined #openstack-horizon | 16:13 | |
| *** dnguyen has quit IRC | 16:18 | |
| *** ircuser-1 has joined #openstack-horizon | 16:20 | |
| *** cshen has quit IRC | 16:25 | |
| *** e0ne has quit IRC | 16:26 | |
| *** cshen has joined #openstack-horizon | 16:41 | |
| *** cshen has quit IRC | 16:47 | |
| *** cshen has joined #openstack-horizon | 16:51 | |
| *** cshen has quit IRC | 16:56 | |
| *** belmoreira has joined #openstack-horizon | 17:02 | |
| *** belmoreira has quit IRC | 17:19 | |
| *** belmoreira has joined #openstack-horizon | 17:20 | |
| *** e0ne has joined #openstack-horizon | 17:31 | |
| *** ChanServ sets mode: +o e0ne | 17:31 | |
| *** tolisbar has quit IRC | 17:37 | |
| *** jpich has joined #openstack-horizon | 17:38 | |
| *** e0ne has quit IRC | 17:47 | |
| *** cshen has joined #openstack-horizon | 17:48 | |
| *** e0ne has joined #openstack-horizon | 17:59 | |
| *** ChanServ sets mode: +o e0ne | 17:59 | |
| *** e0ne has quit IRC | 18:02 | |
| *** e0ne has joined #openstack-horizon | 18:05 | |
| *** ChanServ sets mode: +o e0ne | 18:05 | |
| *** e0ne has quit IRC | 18:07 | |
| *** btully has joined #openstack-horizon | 18:08 | |
| *** btully has quit IRC | 18:13 | |
| *** mvkr has quit IRC | 18:47 | |
| *** belmoreira has quit IRC | 18:48 | |
| *** jpich has quit IRC | 18:59 | |
| *** itlinux has quit IRC | 19:01 | |
| *** pcaruana has quit IRC | 19:05 | |
| *** belmoreira has joined #openstack-horizon | 19:32 | |
| *** itlinux has joined #openstack-horizon | 19:39 | |
| *** mvkr has joined #openstack-horizon | 19:51 | |
| openstackgerrit | Andrew Bogott proposed openstack/horizon master: Server overview: display hypervisor name if available https://review.openstack.org/614328 | 19:54 |
| *** cshen has quit IRC | 19:57 | |
| *** btully has joined #openstack-horizon | 19:58 | |
| *** btully has quit IRC | 20:02 | |
| *** cshen has joined #openstack-horizon | 20:09 | |
| *** e0ne has joined #openstack-horizon | 20:11 | |
| *** ChanServ sets mode: +o e0ne | 20:11 | |
| *** belmoreira has quit IRC | 20:12 | |
| *** itlinux has quit IRC | 20:14 | |
| *** itlinux has joined #openstack-horizon | 20:14 | |
| *** e0ne has quit IRC | 20:28 | |
| *** dave-mccowan has quit IRC | 20:32 | |
| *** belmoreira has joined #openstack-horizon | 20:35 | |
| *** belmoreira has quit IRC | 20:35 | |
| *** rdopiera has quit IRC | 20:50 | |
| *** lblanchard has quit IRC | 20:53 | |
| *** btully has joined #openstack-horizon | 21:46 | |
| *** btully has quit IRC | 21:51 | |
| openstackgerrit | Adrian Turjak proposed openstack/horizon master: Fix django.contrib.auth.middleware monkey patching https://review.openstack.org/614673 | 21:53 |
| adriant | mgagne: ^ see if that works for you | 22:14 |
| mgagne | adriant: I didn't get any error after adding the optional parameter to constructor. So I suspect middleware doesn't need to be a callable in Django 1.11. | 22:15 |
| mgagne | let me test | 22:16 |
| mgagne | to make sure | 22:16 |
| *** threestrands has joined #openstack-horizon | 22:17 | |
| mgagne | no issue on my side | 22:17 |
| mgagne | adriant: ^ | 22:18 |
| adriant | cool, just backporting it to my horizon to test | 22:20 |
| adriant | sweet, seems to work | 22:28 |
| *** Leo_m has quit IRC | 22:36 | |
| *** cshen has quit IRC | 22:56 | |
| *** lblanchard has joined #openstack-horizon | 22:58 | |
| *** mvkr has quit IRC | 23:22 | |
| *** mvkr has joined #openstack-horizon | 23:23 | |
| *** btully has joined #openstack-horizon | 23:35 | |
| *** btully has quit IRC | 23:39 | |
| *** gyee has quit IRC | 23:58 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!