14:01:06 #startmeeting glance 14:01:08 Meeting started Thu Sep 3 14:01:06 2020 UTC and is due to finish in 60 minutes. The chair is abhishekk. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:01:09 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:01:11 The meeting name has been set to 'glance' 14:01:12 #topic roll call 14:01:17 #link https://etherpad.openstack.org/p/glance-team-meeting-agenda 14:01:21 o/ 14:01:40 o/ 14:02:07 lets wait couple of minutes for others 14:03:02 o/ 14:03:31 i am only partially here 14:03:38 ack 14:03:42 lets start 14:03:43 #topic Updates 14:04:19 I have created planning etherpad for W PTG which will be between October 26 to 30 14:04:34 #link https://etherpad.opendev.org/p/Glance-Wallaby-PTG-planning 14:04:49 I have added some of the topics which we will discuss during the PTG 14:05:11 Kindly go through it and add any topics you want to discuss or contribute 14:06:07 Also, I will create specs folder for Wallaby so that we can move/add specs for reviews 14:06:24 moving ahead 14:06:36 #topic release/periodic jobs updates 14:06:51 glance-store 2.3.0 released for Victoria 14:07:35 I may need to submit another patch to create stable/victoria branch for glance-store (which I discuss with smcginnis latter) 14:08:02 Victoria milestone 3 is next week 14:08:19 We have couple of important patches left but those are in good shape 14:08:38 We are maintaining the progress/status in below etherpad 14:08:41 #link https://etherpad.opendev.org/p/glance-victoria-milestone-3-todo 14:09:20 Periodic jobs are all green 14:09:30 Moving ahead 14:09:38 #topic Important reviews 14:09:44 Cinder multiple stores support - https://review.opendev.org/748039 14:10:14 dansmith, has given couple of comments related to tests and whoami-rajat__ is working on those 14:10:31 documentation patch for the same is also up and mentioned in above etherpad 14:10:53 Kindly put some time to review those patches 14:10:57 just writing my thoughts on the functional tests in glance channel 14:12:04 dansmith, there is one major concern in writing functional test for the same 14:12:48 what, mocking cinder? 14:13:08 so you are ok with mocking cinder call, right? 14:13:12 of course 14:13:27 ok 14:13:48 whoami-rajat__, do you have any other concerns? 14:14:11 no 14:14:23 cool, may be we can discuss this offline 14:14:44 Moving to open discussion 14:14:45 sure, thanks for the feedback 14:14:53 #topic Open discussion 14:15:21 I am not sure who has added that topic under open discussion 14:15:24 abhishekk: We can do the glance_store stable branch separately. It's easier when it's done with the "final" release for the cycle, but not a requirement. 14:15:25 may be Steap ? 14:15:49 smcginnis, cool, thank you 14:16:35 oh yeah that was me :) 14:16:42 SOrry, forgot to add my name 14:17:06 So I feel like it's not a very urgent topic, but some users complain that we can give gibberish to "--file" instead of an actual image 14:17:07 ack, do you want to spread some lights on it? 14:17:16 and would like some form of validation 14:17:26 which might be done through a plugin 14:17:33 I wonder is that topic has ever come up before? 14:17:46 not that I am aware of 14:18:20 so I understand the need, but I'm not sure exactly how important it is 14:18:26 presumably the only way we would validate, is if they say it's something like a qcow but isn't right? 14:18:33 yeah 14:18:41 raw can be garbage by design 14:18:49 yep, that is the issue 14:18:50 so we merged the image stream inspection code, 14:19:00 the user gave me an example where they did "echo blabla > file.qcow" 14:19:02 which can detect that an image is at least on its face a given format, 14:19:11 and we could make that enforceable 14:19:37 of course, you could have a qcow containing a bunch of garbage 14:19:40 yes 14:19:42 i.e. stop accepting the stream once it gets far enough for us to determine that it's not the format they say it is 14:19:52 I'm afraid that if we say "hey, we validate images", users might get their hopes up 14:19:54 I am also not sure how this will be beneficial or overhead, but we can have a lite spec for the same 14:19:59 so we just need to be clear that we are only validating the format 14:20:18 this could be optional, with a --validate-format option 14:20:20 rosmaita: yeah, this would be more like a "I think you gave us the wrong file" sort of thing to save time and bandwidth and maybe user frustration trying to boot an instance they don't realize won't work 14:20:22 and/or config option 14:20:49 of course they could also write a shell script that does the validation bfore creatingthe image, in a true UNIX fashion 14:21:12 Introducing config option for it doesn't sound good, or we can also do it at client side? 14:21:39 why doesn't it sound good? 14:21:59 don't we already have a "reject images not in format X" option? 14:22:15 which just requires them to claim the image is the right type 14:22:30 having another one that says "...and check that they're honest" seems good to me 14:23:02 the details could be discussed during the next PTG (or the one after that :p) 14:23:25 I would suggest add this topic in PTG etherpad I just shared above 14:23:26 I just wanted to make sure this was not a "hard NO" so that I can get back to a potentially impatient user :p 14:23:31 abhishekk: yes 14:23:33 will do 14:23:42 Cool 14:24:34 Do we have any other topic to discuss? 14:25:39 Just for FYI, Monday 07 September will be US holiday and most of the people will not be around 14:25:50 yup 14:26:00 So I would recommend to review cinder-glance patch between today and tomorrow 14:26:11 we have devstack dependency on that patch as well 14:26:43 I have reno as well config-referesh patch up as well 14:26:54 we should poke qa people about the devstack patch now I think 14:26:54 So we only need release patch now 14:27:12 yes, just poked them before the meeting 14:27:32 cool 14:28:11 great, we are good for today then 14:28:40 lets wrap this up and utilize this time for reviews and other work 14:28:53 thank you all 14:28:56 Thanks! 14:29:23 #endmeeting