*** rosmaita has left #openstack-glance | 00:10 | |
*** gyee has quit IRC | 00:45 | |
*** Liang__ has joined #openstack-glance | 00:52 | |
*** Liang__ is now known as LiangFang | 00:53 | |
*** markvoelker has quit IRC | 00:54 | |
*** markvoelker has joined #openstack-glance | 00:54 | |
*** markvoelker has quit IRC | 00:59 | |
*** sayalilunkad has quit IRC | 01:23 | |
*** bhdn has quit IRC | 01:24 | |
*** sayalilunkad has joined #openstack-glance | 01:28 | |
*** bhdn has joined #openstack-glance | 01:32 | |
*** szaher has quit IRC | 01:51 | |
*** markvoelker has joined #openstack-glance | 01:55 | |
*** markvoelker has quit IRC | 02:46 | |
*** markvoelker has joined #openstack-glance | 02:46 | |
*** lifeless has joined #openstack-glance | 03:10 | |
*** abhishekk has joined #openstack-glance | 03:35 | |
*** udesale has joined #openstack-glance | 04:52 | |
*** rcernin has quit IRC | 06:47 | |
*** dpawlik has joined #openstack-glance | 07:24 | |
*** dpawlik has quit IRC | 07:34 | |
*** brinzhang has joined #openstack-glance | 07:34 | |
*** dpawlik has joined #openstack-glance | 07:38 | |
*** brinzhang has quit IRC | 07:57 | |
*** brinzhang has joined #openstack-glance | 07:58 | |
*** markvoelker has quit IRC | 08:08 | |
*** markvoelker has joined #openstack-glance | 08:08 | |
*** markvoelker has quit IRC | 08:13 | |
*** lpetrut has joined #openstack-glance | 08:15 | |
*** udesale has quit IRC | 08:52 | |
*** LiangFang has quit IRC | 08:53 | |
*** udesale has joined #openstack-glance | 08:54 | |
*** lpetrut has quit IRC | 09:18 | |
*** udesale has quit IRC | 09:39 | |
*** udesale has joined #openstack-glance | 09:40 | |
*** markvoelker has joined #openstack-glance | 10:09 | |
*** abhishekk has quit IRC | 10:18 | |
*** brinzhang has quit IRC | 11:28 | |
*** jokke_ has joined #openstack-glance | 12:08 | |
*** rosmaita has joined #openstack-glance | 12:34 | |
*** lifeless has quit IRC | 12:42 | |
openstackgerrit | Erno Kuvaja proposed openstack/glance master: Change default visibility to 'private' https://review.openstack.org/628430 | 12:43 |
---|---|---|
jokke_ | rosmaita: you around? | 12:46 |
rosmaita | yep | 12:46 |
rosmaita | jokke_: hope you're feeling better today | 12:46 |
jokke_ | I'm very close to explode | 12:46 |
jokke_ | physically better, yes. | 12:47 |
jokke_ | fecking half of our tests are still using is_public | 12:47 |
jokke_ | that and the tooling to work around it is making me crazy trying to get the default visibility changed | 12:48 |
rosmaita | that is a bummer | 12:49 |
jokke_ | I think what I uploaded now to gerrit narrows it down to like 4-5 failing tests | 12:49 |
jokke_ | that I just haven't got figured out how to work around other than throwing half of our tests away and all the BS that is handling the is_public cases | 12:50 |
jokke_ | btw it's not just this visibility change. I think we are currently ensuring that about half of the legacy stuff is totally broken if it's passed visibility instead of is_public ... like some of the context code, replicator and cache | 12:52 |
rosmaita | that sounds ... unpleasant. | 12:54 |
jokke_ | obviously most of it and the utility functions were necessary while we still supported the v1 but now when it's removed and we don't need to do it anymore it seems that internally it was deemed easier to write some utility functions to translate the v2 dicts to v1 instead of fixing the code that was actually consuming it | 12:55 |
jokke_ | and obviously shit starts breaking when it's not anymore just public_or_not | 12:55 |
rosmaita | here's an idea you'll love ... introduce a proxy class that has is_public and does the translation | 12:57 |
jokke_ | I'm quite sure these were few of those cases of "Lets just make this minimum viable and we can iterate from there" and those follow-ups just never happened | 12:58 |
jokke_ | I would even almost agree proxying some of this if it wasn't absolutely everywhere below API to DB that is doing this | 12:59 |
rosmaita | gotcha | 12:59 |
rosmaita | so it was probably to keep the registry api working | 13:00 |
rosmaita | because it hasn't been in the database since folsom or thereabouts | 13:00 |
jokke_ | some of it is related to the registry, but v2 registry is doing it as well, the db api's are handling it, most of the factory classes has some form of dealing with it and then lots of the context, cache etc. code is just using the utility function. Which btw is great, it checks if the is_public is passed and throws away visibility if so | 13:02 |
jokke_ | and removing thta blows up few other things as apparently it's not great idea to have both going :P | 13:02 |
rosmaita | yuck | 13:02 |
jokke_ | oh and most of our testing fixtures can't handle visibility at all | 13:03 |
jokke_ | just those seems to be converted that are explicitly dealing with visibility or members | 13:04 |
rosmaita | oh boy | 13:05 |
jokke_ | was it this cycle we were able to get rid of registry all together? | 13:05 |
rosmaita | i think that was the plan | 13:05 |
jokke_ | I thought so as well, just wasn't sure if we had to hang on the v2 registry still this cycle or not. I will have a look next week how much dumping all registry code and it's testing away will clean up that mess | 13:07 |
rosmaita | just checked, deprecated since Queens, so could be removed in Rocky | 13:10 |
rosmaita | i mean Stein! | 13:10 |
jokke_ | yeah, hopefully that helps | 13:10 |
jokke_ | thanks for checking | 13:10 |
rosmaita | this says stein: http://specs.openstack.org/openstack/glance-specs/specs/queens/approved/glance/deprecate-registry.html | 13:10 |
rosmaita | well, the good thing is it's early enough in the cycle that if we can get the reg out in the next 2 weeks or so, we'll have time to catch any weirdness | 13:12 |
jokke_ | yeah, I think it's worth of tagging beta release if we get that out to see if it affects something that pulls tags | 13:13 |
jokke_ | even we're not tagging milestones | 13:13 |
rosmaita | i agree | 13:13 |
jokke_ | well, I'll look at it Thursday. Now I'm gonna go and battle that fresh snow on the driveway :D | 13:14 |
jokke_ | have a great weekend when you get that far ;) | 13:14 |
rosmaita | thanks, you too -- watch your back shoveling | 13:14 |
jokke_ | will do | 13:15 |
rosmaita | anything in particular you need me to look at today? | 13:16 |
*** markvoelker has quit IRC | 13:33 | |
*** markvoelker has joined #openstack-glance | 13:33 | |
*** mvkr has quit IRC | 13:57 | |
*** mvkr has joined #openstack-glance | 14:12 | |
*** zul has joined #openstack-glance | 14:51 | |
*** udesale has quit IRC | 14:59 | |
*** dpawlik has quit IRC | 15:25 | |
openstackgerrit | zhulingjie proposed openstack/glance master: Update hacking version to latest https://review.openstack.org/628506 | 15:59 |
*** dpawlik has joined #openstack-glance | 16:11 | |
*** dpawlik has quit IRC | 16:15 | |
*** _alastor1 has joined #openstack-glance | 16:16 | |
*** itlinux has joined #openstack-glance | 16:33 | |
*** zul has quit IRC | 17:08 | |
*** guilhermesp has joined #openstack-glance | 17:09 | |
guilhermesp | Hello everyone! Does anyone ever used glance to store images with cinder? | 17:09 |
*** imacdonn has quit IRC | 17:59 | |
*** imacdonn has joined #openstack-glance | 18:00 | |
*** gyee has joined #openstack-glance | 18:04 | |
*** MattMan has quit IRC | 18:09 | |
*** MattMan has joined #openstack-glance | 18:09 | |
*** itlinux_ has joined #openstack-glance | 18:33 | |
*** itlinux has quit IRC | 18:35 | |
*** _alastor1 is now known as _alastor_ | 18:40 | |
*** itlinux_ has quit IRC | 19:40 | |
*** lifeless has joined #openstack-glance | 20:12 | |
*** itlinux has joined #openstack-glance | 20:43 | |
*** itlinux_ has joined #openstack-glance | 21:56 | |
*** itlinux has quit IRC | 21:59 | |
*** itlinux_ has quit IRC | 22:24 | |
*** rosmaita has left #openstack-glance | 23:31 | |
*** openstack has joined #openstack-glance | 23:43 | |
*** ChanServ sets mode: +o openstack | 23:43 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!