| *** ducttape_ has quit IRC | 00:06 | |
| *** btully has joined #openstack-glance | 00:15 | |
| *** sdake has quit IRC | 00:21 | |
| *** sdake has joined #openstack-glance | 00:21 | |
| *** sdake has quit IRC | 00:21 | |
| *** ozialien10 has quit IRC | 00:23 | |
| *** btully has quit IRC | 00:27 | |
| *** jamielennox|away is now known as jamielennox | 00:31 | |
| *** btully has joined #openstack-glance | 00:44 | |
| *** ozialien10 has joined #openstack-glance | 00:55 | |
| *** btully has quit IRC | 00:56 | |
| *** btully has joined #openstack-glance | 01:14 | |
| *** btully has quit IRC | 01:27 | |
| *** btully has joined #openstack-glance | 01:29 | |
| *** btully has quit IRC | 01:42 | |
| *** btully has joined #openstack-glance | 01:42 | |
| *** btully has quit IRC | 01:56 | |
| *** btully has joined #openstack-glance | 01:59 | |
| *** btully has quit IRC | 02:24 | |
| *** btully has joined #openstack-glance | 02:27 | |
| *** buzztroll has quit IRC | 02:34 | |
| *** lin_yang has joined #openstack-glance | 02:36 | |
| *** ducttape_ has joined #openstack-glance | 02:37 | |
| *** buzztroll has joined #openstack-glance | 02:39 | |
| *** buzztroll has quit IRC | 02:39 | |
| *** houming has joined #openstack-glance | 02:39 | |
| *** btully has quit IRC | 02:51 | |
| *** dims_ has quit IRC | 02:55 | |
| *** dims has joined #openstack-glance | 03:00 | |
| *** TravT has joined #openstack-glance | 03:18 | |
| openstackgerrit | Brian Rosmaita proposed openstack/glance: WADL to RST migration (part 1) https://review.openstack.org/312259 | 03:30 |
|---|---|---|
| openstackgerrit | Brian Rosmaita proposed openstack/glance: WIP - WADL to RST migration (part 2) https://review.openstack.org/315335 | 03:30 |
| *** greghaynes has quit IRC | 03:31 | |
| *** btully has joined #openstack-glance | 03:43 | |
| *** ducttape_ has quit IRC | 03:48 | |
| *** btully has quit IRC | 03:55 | |
| *** pdeore has joined #openstack-glance | 04:05 | |
| *** pdeore has quit IRC | 04:12 | |
| *** btully has joined #openstack-glance | 04:12 | |
| *** links has joined #openstack-glance | 04:13 | |
| *** pdeore has joined #openstack-glance | 04:15 | |
| *** btully has quit IRC | 04:25 | |
| *** gjayavelu has joined #openstack-glance | 04:26 | |
| *** buzztroll has joined #openstack-glance | 04:27 | |
| *** sdake_ has joined #openstack-glance | 04:30 | |
| *** buzztroll has quit IRC | 04:32 | |
| *** sudipto has joined #openstack-glance | 04:32 | |
| *** btully has joined #openstack-glance | 04:44 | |
| *** flwang1 has quit IRC | 04:56 | |
| *** houming has quit IRC | 05:15 | |
| *** rchurch has quit IRC | 05:19 | |
| *** buzztroll has joined #openstack-glance | 05:22 | |
| *** houming has joined #openstack-glance | 05:22 | |
| *** buzztroll has quit IRC | 05:22 | |
| *** buzztroll has joined #openstack-glance | 05:22 | |
| *** abhishekk has joined #openstack-glance | 05:31 | |
| *** groen692 has joined #openstack-glance | 05:39 | |
| *** houming has quit IRC | 05:43 | |
| *** rcernin has joined #openstack-glance | 05:48 | |
| *** sudipto has quit IRC | 05:58 | |
| *** greghaynes has joined #openstack-glance | 06:02 | |
| *** belmoreira has joined #openstack-glance | 06:09 | |
| *** abhishekk has quit IRC | 06:22 | |
| *** david-lyle_ has joined #openstack-glance | 06:26 | |
| *** timburke has quit IRC | 06:26 | |
| *** david-lyle has quit IRC | 06:28 | |
| *** boris-42 has quit IRC | 06:28 | |
| *** SamYaple has quit IRC | 06:28 | |
| *** SamYaple has joined #openstack-glance | 06:29 | |
| *** boris-42 has joined #openstack-glance | 06:29 | |
| *** timburke has joined #openstack-glance | 06:32 | |
| *** gjayavelu has quit IRC | 06:37 | |
| openstackgerrit | Kairat Kushaev proposed openstack/glance: [WIP] Small bug fixes for Glare https://review.openstack.org/318708 | 06:41 |
| *** jokke_ has quit IRC | 06:45 | |
| *** jokke_ has joined #openstack-glance | 06:45 | |
| *** sgotliv_ has joined #openstack-glance | 06:52 | |
| *** ihrachys has joined #openstack-glance | 07:19 | |
| *** abhishekk has joined #openstack-glance | 07:24 | |
| abhishekk | jokke_: hi you around? | 07:26 |
| *** ihrachys has quit IRC | 07:28 | |
| *** jamielennox is now known as jamielennox|away | 07:49 | |
| *** mosulica has joined #openstack-glance | 07:51 | |
| *** ihrachys has joined #openstack-glance | 07:59 | |
| *** ihrachys_ has joined #openstack-glance | 08:02 | |
| *** daemontool has joined #openstack-glance | 08:03 | |
| *** ihrachy__ has joined #openstack-glance | 08:04 | |
| *** ihrachys has quit IRC | 08:05 | |
| *** ihrachy__ has quit IRC | 08:05 | |
| openstackgerrit | Sharat Sharma proposed openstack/python-glanceclient: Removed the deprecated options from the help message https://review.openstack.org/319104 | 08:06 |
| *** ihrachys_ has quit IRC | 08:07 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/glance: [WIP] Testing latest u-c https://review.openstack.org/318436 | 08:10 |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/glance: [WIP] Testing latest u-c https://review.openstack.org/318436 | 08:10 |
| *** dshakhray has joined #openstack-glance | 08:11 | |
| *** jistr has joined #openstack-glance | 08:15 | |
| *** btully has quit IRC | 08:26 | |
| *** houming has joined #openstack-glance | 08:34 | |
| *** sudipto has joined #openstack-glance | 08:42 | |
| *** btully has joined #openstack-glance | 08:43 | |
| *** gjayavelu has joined #openstack-glance | 08:49 | |
| *** btully has quit IRC | 08:52 | |
| *** gjayavelu has quit IRC | 08:54 | |
| *** buzztroll has quit IRC | 08:59 | |
| *** flwang1 has joined #openstack-glance | 09:00 | |
| *** daemontool has quit IRC | 09:05 | |
| *** daemontool has joined #openstack-glance | 09:05 | |
| *** e0ne has joined #openstack-glance | 09:06 | |
| *** dmk0202 has joined #openstack-glance | 09:19 | |
| *** mvk has joined #openstack-glance | 09:21 | |
| openstackgerrit | Darja Shakhray proposed openstack/glance: [WIP] Update filters for Glare https://review.openstack.org/319155 | 09:27 |
| *** houming has quit IRC | 09:40 | |
| openstackgerrit | Darja Shakhray proposed openstack/glance: [WIP] Update filters for Glare https://review.openstack.org/319155 | 09:45 |
| *** btully has joined #openstack-glance | 09:49 | |
| *** btully has quit IRC | 10:01 | |
| *** buzztroll has joined #openstack-glance | 10:14 | |
| *** ihrachys has joined #openstack-glance | 10:15 | |
| *** buzztroll has quit IRC | 10:18 | |
| *** ihrachys has quit IRC | 10:20 | |
| *** mosulica has quit IRC | 10:24 | |
| *** ihrachys has joined #openstack-glance | 10:25 | |
| *** ihrachys has quit IRC | 10:31 | |
| openstackgerrit | Sharat Sharma proposed openstack/python-glanceclient: Removed the --no-ssl-compression parameter which is deprecated https://review.openstack.org/319189 | 10:34 |
| *** btully has joined #openstack-glance | 10:42 | |
| *** btully has quit IRC | 10:47 | |
| *** btully has joined #openstack-glance | 10:48 | |
| *** btully has quit IRC | 11:01 | |
| *** sathiya has joined #openstack-glance | 11:05 | |
| sathiya | Hi, all | 11:05 |
| sathiya | Is there any difference between "nova image-list" and "glance image-list" ? | 11:05 |
| sathiya | Thanks in Advance. | 11:06 |
| kragniz | sathiya: nova image-list is a proxy for glance's image-list | 11:06 |
| sudipto | sathiya, the nova image-list calls into glance to show the image-lists - however it currently defaults to using glance V1 APIs as opposed to glance which currently defaults to V2 | 11:07 |
| sathiya | Then why we have two commads? | 11:07 |
| sudipto | sathiya, because traditionally everything came out of Nova i would say :) | 11:08 |
| johnthetubaguy | its old history, glance v1 was never exposed, so Nova had to create a proxy | 11:08 |
| johnthetubaguy | basically, use glance image-list | 11:08 |
| sudipto | johnthetubaguy, +1 :D | 11:08 |
| johnthetubaguy | nova image-list is a historic thing for backwards compatiblity | 11:08 |
| sathiya | How can we test this ? Any simple example might be help us ? | 11:12 |
| johnthetubaguy | you can just use those cli calls you posted | 11:15 |
| sathiya | Johnthetubaguy: krangniz: sudipto: Thanks a lot :) | 11:17 |
| *** mfedosin has joined #openstack-glance | 11:22 | |
| *** mfedosin has quit IRC | 11:27 | |
| sathiya | Hi, all | 11:32 |
| sathiya | In which version Glance component introduced in openstack | 11:32 |
| sathiya | please someone help on this | 11:33 |
| sathiya | Any help Highly appreciated | 11:33 |
| *** daemontool has quit IRC | 11:34 | |
| *** sathiya has left #openstack-glance | 11:36 | |
| *** audit has joined #openstack-glance | 11:40 | |
| rosmaita | sathiya: Glance was introduced in Bexar | 11:41 |
| *** daemontool has joined #openstack-glance | 11:41 | |
| *** mfedosin has joined #openstack-glance | 11:42 | |
| *** smatzek has joined #openstack-glance | 11:45 | |
| *** ihrachys has joined #openstack-glance | 11:49 | |
| *** btully has joined #openstack-glance | 11:53 | |
| openstackgerrit | Darja Shakhray proposed openstack/glance-specs: Add transaction layer https://review.openstack.org/315483 | 11:54 |
| *** raildo-afk is now known as raildo | 11:55 | |
| *** audit has quit IRC | 11:58 | |
| *** ducttape_ has joined #openstack-glance | 11:59 | |
| *** sudipto has quit IRC | 12:00 | |
| openstackgerrit | Darja Shakhray proposed openstack/glance-specs: Add transaction layer https://review.openstack.org/315483 | 12:02 |
| *** julim has joined #openstack-glance | 12:02 | |
| *** buzztroll has joined #openstack-glance | 12:02 | |
| *** flwang has quit IRC | 12:03 | |
| *** sathiya has joined #openstack-glance | 12:03 | |
| sathiya | rosmaita: Thankyou. | 12:03 |
| rosmaita | sathiya: np. here's a quick "history of glance": http://www.slideshare.net/racker_br/the-evolution-of-glance-api-on-the-way-from-v1-to-v3 | 12:04 |
| *** btully has quit IRC | 12:05 | |
| *** buzztroll has quit IRC | 12:07 | |
| *** flwang has joined #openstack-glance | 12:11 | |
| *** daemontool_ has joined #openstack-glance | 12:14 | |
| *** sgotliv_ has quit IRC | 12:14 | |
| *** ducttape_ has quit IRC | 12:15 | |
| *** ayoung has quit IRC | 12:16 | |
| *** daemontool has quit IRC | 12:17 | |
| *** user154752 has joined #openstack-glance | 12:30 | |
| rosmaita | jokke_: very nice response on that email thread! | 12:39 |
| *** pdeore has quit IRC | 12:43 | |
| *** marcusvrn_ has joined #openstack-glance | 12:47 | |
| *** sathiya has quit IRC | 12:47 | |
| *** julim has quit IRC | 12:49 | |
| *** btully has joined #openstack-glance | 12:54 | |
| jokke_ | rosmaita: cheers ... you see I _can_ be nice, polite and helpful ;) | 12:54 |
| rosmaita | jokke_: *I* know that! we just need to get the message out to the rest of the team! :) | 12:55 |
| rosmaita | jokke_: but seriously, that was very nicely done | 12:55 |
| *** julim has joined #openstack-glance | 12:57 | |
| jokke_ | :) | 12:59 |
| *** buzztroll has joined #openstack-glance | 13:00 | |
| *** buzztroll has quit IRC | 13:05 | |
| *** btully has quit IRC | 13:06 | |
| *** daemontool_ has quit IRC | 13:09 | |
| *** ninag has joined #openstack-glance | 13:28 | |
| *** smatzek has quit IRC | 13:32 | |
| *** chuck__ is now known as zul | 13:33 | |
| *** ayoung has joined #openstack-glance | 13:33 | |
| *** btully has joined #openstack-glance | 13:42 | |
| *** daemontool has joined #openstack-glance | 13:44 | |
| *** links has quit IRC | 13:46 | |
| *** btully has quit IRC | 13:47 | |
| *** ducttape_ has joined #openstack-glance | 13:50 | |
| *** ametts has joined #openstack-glance | 13:53 | |
| nikhil | rosmaita: jokke_ : indeed, what a tremendous response there! kudos! \o \o/ o/ | 13:53 |
| *** btully has joined #openstack-glance | 13:53 | |
| *** smatzek has joined #openstack-glance | 13:55 | |
| *** ninag has quit IRC | 14:01 | |
| *** ninag has joined #openstack-glance | 14:02 | |
| *** btully has quit IRC | 14:05 | |
| *** daemontool has quit IRC | 14:06 | |
| *** ninag has quit IRC | 14:07 | |
| *** pushkaru has joined #openstack-glance | 14:08 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 14:09 | |
| openstackgerrit | Sven Anderson proposed openstack/glance: Let setup.py compile_catalog process all language files https://review.openstack.org/319269 | 14:10 |
| *** btully has joined #openstack-glance | 14:13 | |
| *** hemanthm|afk is now known as hemanthm | 14:14 | |
| *** ozialien10 has quit IRC | 14:14 | |
| nikhil | rosmaita: got a min to talk about the standardization of properties esp in nova? | 14:15 |
| nikhil | rosmaita: reading up some logs, seems like rackspace ops want to do some filtering on deactivated images | 14:15 |
| nikhil | rosmaita: want to make sure we do it right | 14:15 |
| nikhil | apparently glance has many leaks! | 14:15 |
| rosmaita | nikhil: sure, let's talk | 14:16 |
| rosmaita | hadn't heard about the deactivated image filtering | 14:17 |
| nikhil | rosmaita: getting you a link | 14:17 |
| *** btully has quit IRC | 14:18 | |
| nikhil | rosmaita: http://eavesdrop.openstack.org/irclogs/%23openstack-operators/%23openstack-operators.2016-05-20.log.html#t2016-05-20T03:12:14 | 14:18 |
| nikhil | rosmaita: http://eavesdrop.openstack.org/irclogs/%23openstack-operators/%23openstack-operators.2016-05-20.log.html#t2016-05-20T03:18:39 | 14:18 |
| nikhil | rosmaita: http://eavesdrop.openstack.org/irclogs/%23openstack-operators/%23openstack-operators.2016-05-20.log.html#t2016-05-20T03:21:55 | 14:18 |
| nikhil | hemanthm: ^ | 14:19 |
| nikhil | we may need to push a bit on standardizing props | 14:19 |
| nikhil | (Thanks a ton to flwang for the same) | 14:20 |
| nikhil | rosmaita: just read "oh yeah - forgot that Brian was back in the Glance game" that's funny | 14:21 |
| nikhil | lol | 14:21 |
| rosmaita | just saw the same line ... glad to be back in the game! | 14:24 |
| hemanthm | welcome rosmaita | 14:25 |
| jokke_ | nikhil & rosmaita: what do you think about providing the copy from in client for now to address at least some of the use cases, so consumers like deployers could create those base images they need with something like glance image-create ... --download-from https://10.0.0.1/static/base-img.qcow2 | 14:25 |
| rosmaita | there seem to be several different use cases discussed there | 14:26 |
| jokke_ | so that the client would download the image from the URI and upload it to Glance | 14:26 |
| jokke_ | oh hemanthm as well^^ | 14:26 |
| nikhil | jokke_: once it's in the client, it's not going away .. ever! ;-) | 14:26 |
| *** sgotliv_ has joined #openstack-glance | 14:26 | |
| jokke_ | nikhil: true | 14:26 |
| rosmaita | jokke_: that seems to be exactly the "bad experience" you described in your email! | 14:27 |
| jokke_ | nikhil: it's also quite simple bit of code that should be fairly easy to maintain | 14:27 |
| jokke_ | rosmaita: why? | 14:27 |
| rosmaita | jokke_: the problem is with copy-from, i issue a command and glance does the copying, i can close my shell and go home | 14:27 |
| nikhil | I dunno how I feel about the behind the scnes magic on the images | 14:27 |
| rosmaita | with your proposal, i need to hang around as my client downloads and then uploads the image | 14:28 |
| jokke_ | rosmaita: kick it on screen, close your connection and go home | 14:28 |
| nikhil | well I think they can use a simple wget on heat side if they need to | 14:29 |
| rosmaita | jokke_: we could just document the bash commands ... why build it into the client? | 14:29 |
| nikhil | rosmaita: +1 | 14:29 |
| jokke_ | rosmaita: the point is we won't have that existing in the import process for next year give or take | 14:29 |
| nikhil | :-) | 14:30 |
| nikhil | jokke_: exactly my point on ML :D | 14:30 |
| rosmaita | nikhil: flwang: about the image metadata stuff ... i had quasi-volunteered to get a user story going about that | 14:30 |
| rosmaita | haven't had time to follow up yet | 14:30 |
| jokke_ | aand if people are serious about OSC we don't need to carry it on forever after all ;) | 14:30 |
| rosmaita | i think we need to iron out the various use cases | 14:30 |
| rosmaita | jokke_: go OSC FTW!!! | 14:31 |
| nikhil | jokke_ there are stakeholders for all things but no one is willing to commit to certain things. if things are not there, it will become priority and people who have conflicting opinions they will make sure to show up and get things done :) | 14:31 |
| jokke_ | rosmaita: yup, then we don't need to worry about the CLI anymore | 14:31 |
| nikhil | well | 14:31 |
| nikhil | that's not true | 14:31 |
| nikhil | CLI can't go away for at least 2 years after deprecation is what I was told yesterday | 14:32 |
| nikhil | client side stuff is way way slower than we expect | 14:32 |
| nikhil | people get comfortable and then they do not want to switch | 14:32 |
| jokke_ | nikhil: absolutely is ... as soon as we drop CLI support from python-glanceclient I'm not gonna worry about Glance CLI and if someone gets amazing idea to use OS SDK instead of python-glanceclient I'm not gonna worry about the client either :P | 14:33 |
| jokke_ | not projects in my portfolio :P | 14:33 |
| jokke_ | nikhil: that's great news | 14:33 |
| nikhil | jokke_: you saw the dis-satisfaction people have when things break | 14:34 |
| nikhil | even client side | 14:34 |
| nikhil | who don't contribute upstream!!!! | 14:34 |
| jokke_ | nikhil: yup | 14:34 |
| nikhil | I am planning to ask this question on the "what is openstack" thread today (nicely) | 14:34 |
| jokke_ | nikhil: I think it was amazing how small it was when we moved to 1.0 compared the user base | 14:34 |
| nikhil | :) | 14:34 |
| jokke_ | few crying how their 3 year old shell scripts broke ... buuhuhuu go and fix your scripts | 14:35 |
| rosmaita | jokke_: +1 | 14:35 |
| nikhil | jokke_: lol | 14:35 |
| nikhil | rofl | 14:35 |
| * nikhil still on the floor laughing :D | 14:35 | |
| nikhil | have mob in my hand :D | 14:36 |
| jokke_ | rosmaita: back to your question why just not document how to do it in bash, that comes to the usability again ... do I need to fire one command to get it done or do I need wrap my client around some bash/python/perl/pick-your-fav script | 14:36 |
| rosmaita | i know, but you've already included the fire-up-screen step ... which i don't know is available to windows users? | 14:37 |
| nikhil | rosmaita: lol | 14:37 |
| jokke_ | rosmaita: that same question applies why did we include the upload option to image-create instead of documenting how to do it in bash | 14:37 |
| nikhil | rosmaita: I think we also want to think about rasberry-pi from today onward | 14:37 |
| jokke_ | rosmaita: simple, don't use windows :P | 14:38 |
| jokke_ | I don't think we officially support windows anyways, do we? :P | 14:38 |
| rosmaita | jokke_: that would definitely improve user experience! | 14:38 |
| nikhil | rosmaita: I would agree :) | 14:38 |
| jokke_ | rosmaita: how do you pipe the image to glanceclient in windows? | 14:38 |
| nikhil | microsoft is planning to support linux commands | 14:39 |
| nikhil | we'd just wait for that | 14:39 |
| rosmaita | don't know, i don't use windows! but i hear that some people do | 14:39 |
| nikhil | and not do anything till then :) | 14:39 |
| jokke_ | rosmaita: absolutely their problem :D | 14:39 |
| jokke_ | rosmaita: and they see CLI as bas user experience anyways, they just want to click cute buttons | 14:40 |
| jokke_ | s/bas/bad/ | 14:40 |
| nikhil | https://msdn.microsoft.com/en-us/commandline/wsl/about | 14:40 |
| nikhil | http://www.linux-magazine.com/Online/News/Microsoft-Brings-the-Linux-Command-Line-to-Windows | 14:40 |
| nikhil | our proposal for wget is cross platform | 14:40 |
| nikhil | let's document that! | 14:41 |
| jokke_ | nikhil: I'm pretty sure that has been the great influencer why OSC and OpenStack clients in general has been looking like powershell commands lately | 14:41 |
| jokke_ | Conspiracy!!! | 14:41 |
| nikhil | :) | 14:41 |
| jokke_ | there's some insider who has been planning this all along | 14:42 |
| nikhil | indeed | 14:42 |
| nikhil | it's our fav bill | 14:42 |
| nikhil | :) | 14:42 |
| nikhil | rosmaita: ref: the metadata, we'd just wait? | 14:43 |
| nikhil | rosmaita: I'm assuming there're things I just don't know or you pretty elaborate session proposal covers it all? | 14:43 |
| nikhil | your* | 14:43 |
| *** daemontool has joined #openstack-glance | 14:44 | |
| jokke_ | but seriously doing the download/upload in client would solve those issues for use cases like Heat that don't want to let users access/save data on the systems the client is ran and for those who just want the ease of use | 14:44 |
| jokke_ | and no, this is not solving the usecase for those who want to run their glanceclient locally and are connected to internet via dialup | 14:45 |
| rosmaita | nikhil: i will follow up with flwang (i guess by email, i think it is midnight in his TZ) | 14:46 |
| nikhil | rosmaita: surely, more like 3 am :) | 14:46 |
| *** btully has joined #openstack-glance | 14:46 | |
| jokke_ | so call him :D | 14:46 |
| nikhil | jokke_: please go back to your usualy self :) | 14:46 |
| nikhil | usual* | 14:46 |
| jokke_ | Viva la sunchrounous communication!!! | 14:47 |
| jokke_ | sorry! | 14:47 |
| nikhil | lol | 14:47 |
| nikhil | that's quite funny | 14:47 |
| nikhil | jokke_ I think the softness in your email tone grew on you | 14:47 |
| *** mtanino has joined #openstack-glance | 14:48 | |
| jokke_ | nikhil: end users have always been the soft spot for me ... the ones I care so much less is developers. They can and should adjust or they can just go and do something else :P | 14:49 |
| jokke_ | it's like "But AWS API does this that and monitors your coffee brewer on the side" "Cool, what are you still doing here then" :P | 14:50 |
| nikhil | jokke_: lol | 14:51 |
| nikhil | jokke_: you are on a roll today | 14:51 |
| nikhil | :P | 14:51 |
| nikhil | "monitors your coffee brewer on the side" | 14:52 |
| nikhil | that's the troll post for today | 14:52 |
| nikhil | may you'd tweet it | 14:52 |
| rosmaita | what happened to the coffee-from functionality that was in glance v1 ? | 14:52 |
| *** sudipto has joined #openstack-glance | 14:52 | |
| nikhil | rosmaita: that trumped everything | 14:52 |
| nikhil | !!! | 14:52 |
| openstack | nikhil: Error: "!!" is not a valid command. | 14:52 |
| nikhil | lol | 14:52 |
| nikhil | wondering how I will get through the day w/o laughing :D | 14:53 |
| jokke_ | I got home last night about half two after fixing my car at the motorway resting are in rain and I couldn't get sleep on the night before, so it might have something to do about this mental state | 14:53 |
| nikhil | rosmaita: I am gunna def tweet that! | 14:54 |
| jokke_ | it's great moment driving home and realizing that the engine is running between 114-117C | 14:54 |
| *** btully has quit IRC | 14:55 | |
| nikhil | oh boy | 14:55 |
| jokke_ | so I dug out the thermostat, slaughtered the intestines out of it and topped up the coolant with sparkling water as that was all I got | 14:55 |
| *** btully has joined #openstack-glance | 14:56 | |
| jokke_ | benefit of growing up with truckers and formerly being mechanic at racing team ... it would have been expensive night otherwise | 14:56 |
| rosmaita | jokke_: i am sure your engine appreciated the sparkling water | 14:59 |
| sudipto | (without knowing the context - since i just logged in) - yesterday it rained heavily here in Bangalore as well. So many cars had a break down...Was returning home around 11:30 after a game of football. | 15:00 |
| sudipto | scary as hell! | 15:00 |
| nikhil | jokke_: racing team work sounds interesting. why you work in IT again? | 15:01 |
| jokke_ | rosmaita: on those temps the bubbles will be gone in no time (try to have open bottle of sparkling water on that heat for a moment) ... other than that the salts and minerals in those are in so small amounts that they won't do harm (I had like 60miles to home) ... It difinitely appreciated it more than running without that water | 15:01 |
| jokke_ | nikhil: that was just a hobby | 15:01 |
| jokke_ | nikhil: had to do IT to afford it :D | 15:02 |
| rosmaita | jokke_: i was just kidding, i meant that your engine would appreciate the thoughtfulness of Perrier as opposed to tap water, as a measure of your esteem | 15:02 |
| nikhil | ;) | 15:02 |
| *** dillaman has joined #openstack-glance | 15:02 | |
| jokke_ | rosmaita: refreshing :D | 15:02 |
| jokke_ | Had can of Red Bull as well but my car is on extremely low carb diet | 15:03 |
| rosmaita | :) | 15:03 |
| *** ninag has joined #openstack-glance | 15:03 | |
| nikhil | lol | 15:04 |
| *** ninag has quit IRC | 15:04 | |
| jokke_ | so of my friends have been laughing why I always have my toolbox in the car ... on nights like last it seriously pays off to have tools to do simple task like that on the road | 15:06 |
| *** ninag has joined #openstack-glance | 15:06 | |
| openstackgerrit | Mike Fedosin proposed openstack/glance-specs: Glare API https://review.openstack.org/283136 | 15:06 |
| jokke_ | s/so/some/ | 15:06 |
| *** ninag has quit IRC | 15:08 | |
| *** ninag_ has joined #openstack-glance | 15:09 | |
| *** btully has quit IRC | 15:09 | |
| nikhil | :) | 15:11 |
| nikhil | rosmaita: I am a bit worried about this https://etherpad.openstack.org/p/AUS-ops-informal-meetup | 15:15 |
| nikhil | and I dunno how to answer some people who thing projects like to work on vaccuum | 15:16 |
| nikhil | think* | 15:16 |
| nikhil | it is merely where we are looking that is defining what vacuum is | 15:16 |
| nikhil | (anyway, if someone wants to know more on what I mean to say by that I can help answer) | 15:16 |
| *** user154752 has quit IRC | 15:19 | |
| *** sjmc7 has joined #openstack-glance | 15:24 | |
| jokke_ | nikhil: please do ... I'd like to hear your concerns around that | 15:25 |
| rosmaita | nikhil: lot of stuff on that etherpad | 15:25 |
| jokke_ | and one funny thing I picked from that meetup EP ... "Getting the Glance APIs out of path for data transfer seems to be a common desire+1" and same time looking their multizone installations are all running with shared glance | 15:26 |
| *** belmoreira has quit IRC | 15:28 | |
| *** daemontool_ has joined #openstack-glance | 15:31 | |
| *** daemontool has quit IRC | 15:33 | |
| *** btully has joined #openstack-glance | 15:34 | |
| *** sudipto has quit IRC | 15:44 | |
| *** btully has quit IRC | 15:45 | |
| *** rcernin has quit IRC | 15:51 | |
| *** MattMan has quit IRC | 15:56 | |
| *** MattMan has joined #openstack-glance | 15:56 | |
| *** djkonro has joined #openstack-glance | 15:57 | |
| *** sdake_ has quit IRC | 15:59 | |
| *** btully has joined #openstack-glance | 16:00 | |
| *** sdake has joined #openstack-glance | 16:01 | |
| *** sjmc7 has left #openstack-glance | 16:02 | |
| tsymanczyk | hello. re the community image sync up there isn't much to say from my end. there are still open questions on the spec but as far as i'm aware nobody is waiting on my input. i've been trying to get the coding started but some unrelated things at work have not allowed me to progress as quickly as i would prefer. hoping to get a first minimal patchset soon. | 16:02 |
| tsymanczyk | i assume people like seeing progress. | 16:02 |
| rosmaita | tsymanczyk: hello! am looking at the spec now (got delayed leaving my comments) | 16:02 |
| rosmaita | was there still a question about what to do with the member list if you do a shared -> private visibility change? | 16:03 |
| tsymanczyk | i think that whole area is still an open question. | 16:04 |
| rosmaita | ok | 16:04 |
| rosmaita | my feeling is we should leave it, unless doing so is going to cause complications (i don't know, maybe database complexity or something) | 16:05 |
| rosmaita | i mean, leave the member-list on the image | 16:05 |
| tsymanczyk | i think that's the way it is in both the bp and spec. sounds good to me. | 16:06 |
| *** e0ne has quit IRC | 16:06 | |
| tsymanczyk | there are two big open questions that i'm aware of, aside from (additionally) the "hard cases" listed in one of the etherpads. | 16:07 |
| *** daemontool has joined #openstack-glance | 16:07 | |
| tsymanczyk | one being how to strike the right balance between this being a self-service thing (as you had pointed out in a review comment) while still maintaining anti-spam properties (as nikhil had pointed out). he and i spoke at length last friday, and the summary of that should be in the review comments. i know he wanted to wait for your input. | 16:08 |
| *** daemontool_ has quit IRC | 16:09 | |
| tsymanczyk | the other open question being the topic you brought up re line :377 | 16:09 |
| tsymanczyk | there are a lot of potential ways to sanely do this, and i don't know how we decide which is "right". | 16:09 |
| tsymanczyk | i don't have an opinion that i feel i need to push. i only care that it's sane and "right" for the community. | 16:10 |
| rosmaita | i was just putting a comment at line 377 that i agree with your analysis, namely, that trying for backwards compatability will just make it harder to understand how to use the feature | 16:11 |
| rosmaita | but i don't know what our obligations are | 16:12 |
| *** sdake has quit IRC | 16:12 | |
| rosmaita | may need to talk to some api-wg people | 16:13 |
| rosmaita | maybe sigmavirus24 | 16:13 |
| *** dmk0202 has quit IRC | 16:17 | |
| *** dshakhray has quit IRC | 16:18 | |
| sigmavirus24 | rosmaita: I'm not sure what you'er talking about specifically | 16:22 |
| openstackgerrit | Sven Anderson proposed openstack/glance: Let setup.py compile_catalog process all language files https://review.openstack.org/319269 | 16:23 |
| nikhil | (sorry about the delay, got pulled into a urgent stuff) | 16:25 |
| rosmaita | sigmavirus24: can you take a look at https://review.openstack.org/#/c/271019/17/specs/newton/approved/glance/community_visibility.rst | 16:25 |
| nikhil | (I've been reading and followin along the convo) | 16:25 |
| rosmaita | specifically, the part that starts at line 360 | 16:25 |
| rosmaita | sigmavirus24: ^^ | 16:25 |
| nikhil | rosmaita: I do want to discuss the necessity for keeping member-list on a private image. what's it req. for? | 16:27 |
| rosmaita | nikhil: gimme a sec, was just writing that on the spec | 16:27 |
| *** daemontool has quit IRC | 16:27 | |
| *** daemontool_ has joined #openstack-glance | 16:27 | |
| nikhil | kk | 16:28 |
| sigmavirus24 | rosmaita: so notifications aren't covered by teh API-WG last I was involved, but that very well may have changed | 16:28 |
| sigmavirus24 | rosmaita: that said, notifications are some sort of contract | 16:28 |
| rosmaita | sigmavirus24: i must have given you the wrong line # | 16:28 |
| nikhil | but he brought a good point too! | 16:28 |
| sigmavirus24 | And I think more than Searchlight relies on our notifications (i.e., ceilometer might) | 16:28 |
| nikhil | nova is planning on doing versioned notifications | 16:28 |
| rosmaita | I meant the "other end user impact" section | 16:29 |
| sigmavirus24 | So keeping is_public and adding visibility makes sense | 16:29 |
| sigmavirus24 | rosmaita: eh, sorry, I saw comments (in gertty) that were around line 360 and read those :() | 16:29 |
| rosmaita | sigmavirus24: np, your comment about notifications is consistent with what we were thinking, so that's good to know | 16:30 |
| sigmavirus24 | rosmaita: uh, so I'm opposed to the idea of an owner having to set it to shared | 16:30 |
| sigmavirus24 | in particular, I think I can keep an image private and still add a new tenant, can't I/ | 16:30 |
| nikhil | today you can | 16:31 |
| sigmavirus24 | I'm also fairly certain they can see it without a problem at that point too, regardless of visibility | 16:31 |
| sigmavirus24 | nikhil: right | 16:31 |
| rosmaita | sigmavirus24: right, but we are proposing to change that | 16:31 |
| sigmavirus24 | rosmaita: I get that | 16:31 |
| sigmavirus24 | And don't quite like it | 16:31 |
| rosmaita | hmmm | 16:32 |
| rosmaita | well, we want the behavior to be consistent across all "types" of sharing | 16:32 |
| sigmavirus24 | So, perhaps this is a stupid question | 16:33 |
| sigmavirus24 | (or a couple stupid questions) | 16:33 |
| sigmavirus24 | We're going to start tracking member status on an image? i.e., it will no longer be "member" or "not a member" | 16:33 |
| nikhil | same as locations :) | 16:33 |
| sigmavirus24 | How can a "shared" image that is not public have a Pending member? | 16:34 |
| sigmavirus24 | nikhil: I'm not sure what you mean | 16:34 |
| sigmavirus24 | So here's my concern | 16:35 |
| sigmavirus24 | Public and Private already exist | 16:35 |
| nikhil | sigmavirus24: I like your questions. I was stating that we are going in the similar direction as locations (locs- in particular are something I don't like) but for members this makes sense. | 16:35 |
| tsymanczyk | i think the idea is for there to be three different statuses for members. pending (invited and not responded), rejected, and approved. | 16:35 |
| sigmavirus24 | tsymanczyk: how do members get notified that they have been invited? | 16:35 |
| *** buzztroll has joined #openstack-glance | 16:35 | |
| tsymanczyk | good question. | 16:35 |
| sigmavirus24 | do other services have a similar concept of "inviting" a tenant to share something? | 16:36 |
| rosmaita | sigmavirus24: you don't invite the tenant | 16:36 |
| rosmaita | the member_status is the *member*'s attitude toward what you shared | 16:36 |
| sigmavirus24 | rosmaita: so that means that someone can somehow get an image-id from somewhere and petition to become a member? | 16:36 |
| rosmaita | yep, if they also know how to contact you | 16:36 |
| nikhil | sigmavirus24: shared images are discoverable | 16:36 |
| nikhil | even if not accepted | 16:36 |
| rosmaita | nikhil: only with the people they are shared with, though | 16:37 |
| sigmavirus24 | rosmaita: that's so very confusing | 16:37 |
| nikhil | rosmaita: correct, thanks for stating that explicitly | 16:37 |
| rosmaita | sigmavirus24: from the producer side, when i make someone a member, that's it, they are a member | 16:37 |
| sigmavirus24 | rosmaita: right | 16:37 |
| rosmaita | but from the consumer side, they decide whether they want to "see" the image in their image-list or not | 16:38 |
| rosmaita | that's all the member_status does | 16:38 |
| sigmavirus24 | ah | 16:38 |
| nikhil | sigmavirus24: this is a trust based sharing model. you only share with someone you know. for community it's different, it's open sharing without contract of a public image. | 16:38 |
| sigmavirus24 | that name and the states for it are terribly confusing | 16:38 |
| sigmavirus24 | nikhil: I'm totally on board with community sharing :D | 16:38 |
| nikhil | :) | 16:38 |
| sigmavirus24 | It's this new shared visibility model that I'm having trouble wrapping my head around | 16:38 |
| nikhil | sigmavirus24: the idea is to make sharing more explicit | 16:39 |
| rosmaita | well, it's the same, except that the image must *also* have visibility == shared for the sharing to work | 16:39 |
| nikhil | sigmavirus24: currently, visibility is implied if shared (no DB storage or transactions associated with it) | 16:39 |
| sigmavirus24 | nikhil: is it? | 16:39 |
| sigmavirus24 | Today, the visibility of my image has no impact on who I can add to the image as a member | 16:40 |
| sigmavirus24 | Let me make a table of how I think this might work better that's still somewhat backwards compatible | 16:41 |
| rosmaita | and it still won't, you can add anyone you like, the image won't be "published" until you make visibility == shared | 16:41 |
| sigmavirus24 | So the list of transitions on L164 is wrong then? | 16:42 |
| rosmaita | and if you make visibility == community, that image will still be "bookmarked" for all the current members, so it won't disappear on them | 16:42 |
| sigmavirus24 | Why does private -> public erase image members? | 16:42 |
| sigmavirus24 | Maybe I'm wrong, but can't I currently add image members to a public image? | 16:42 |
| rosmaita | no point having them on a public image | 16:42 |
| rosmaita | no, you can't | 16:43 |
| rosmaita | get an error | 16:43 |
| rosmaita | a real one, not a 500 | 16:43 |
| nikhil | sigmavirus24: https://github.com/openstack/glance/blob/master/glance/db/__init__.py#L189 | 16:43 |
| sigmavirus24 | rosmaita: https://github.com/openstack/glance/blob/master/glance/db/__init__.py#L339 doesn't seem to care, so it's enforced at a higher level? | 16:45 |
| *** djkonro has quit IRC | 16:46 | |
| rosmaita | sigmavirus24: guess so ... i'm working from memory, though | 16:47 |
| nikhil | sigmavirus24: I don't think you can add member to an image that you don't own | 16:47 |
| nikhil | sigmavirus24: so, public images are not owned by you | 16:47 |
| rosmaita | i'm fairly certain i've seen an error message saying you can't put members on a public image | 16:47 |
| rosmaita | actually, i saw it about a month ago when we had a use case for it | 16:48 |
| nikhil | I will need to dig the authz proxy | 16:48 |
| sigmavirus24 | nikhil: ah I think that's what I was missing | 16:48 |
| nikhil | k, found it | 16:48 |
| nikhil | https://github.com/openstack/glance/blob/master/glance/api/authorization.py#L127 | 16:49 |
| nikhil | it's not really top level | 16:50 |
| nikhil | but again one of the hacks, I really dislike in our system | 16:50 |
| *** pdeore has joined #openstack-glance | 16:51 | |
| sigmavirus24 | So https://github.com/openstack/glance/blob/da31cfff614b46aaf68edf25daf85b9b464f449d/glance/db/sqlalchemy/api.py#L780 will raise an error if the iamge is public and not mutable | 16:51 |
| sigmavirus24 | That's in image update though | 16:51 |
| sigmavirus24 | nikhil: thanks for that | 16:52 |
| nikhil | so for image members, this is where it should fail | 16:52 |
| nikhil | https://github.com/openstack/glance/blob/master/glance/api/authorization.py#L163 | 16:52 |
| sigmavirus24 | So has anyone round-tripped an image recently to see if private + members -> public -> private retains members/ | 16:52 |
| rosmaita | sigmavirus24: i have not | 16:53 |
| nikhil | sigmavirus24: you got a point | 16:53 |
| nikhil | I (intuitively) think it should | 16:53 |
| sigmavirus24 | I can understand disabling modifying the member list for that transition | 16:53 |
| nikhil | as we do not do 'background checks' on the image while making it public | 16:53 |
| sigmavirus24 | but deleting current members seems.... bad | 16:53 |
| nikhil | so this should disallow adding more members https://github.com/openstack/glance/blob/master/glance/api/authorization.py#L163 | 16:54 |
| nikhil | for retention of members, we do not check anything afair | 16:54 |
| nikhil | (background cleanup is mostly upto users and ops) | 16:55 |
| sigmavirus24 | so | 16:55 |
| sigmavirus24 | 1. let's not change that as part of this spec | 16:55 |
| sigmavirus24 | Let's float the idea to operators and do it separately | 16:55 |
| sigmavirus24 | 2. Does this mean that "shared" is a sub-status/visibility of "private" then? | 16:55 |
| sigmavirus24 | because the only way an image can be shared currently is if it's private and has members added to it | 16:56 |
| nikhil | ouch | 16:56 |
| sigmavirus24 | And that's essentially what "shared" will be but with the added ability for users to hide images they don't want to see that are shared with them? | 16:56 |
| sigmavirus24 | er, not the only way it can be shared, but you get what I mean | 16:56 |
| sigmavirus24 | "shared with a limited number of people" | 16:57 |
| tsymanczyk | rosmaita posted a good argument a few minutes ago to the change regarding a case where he wants to retain the the member list on -> private, but still disallow people from seeing the image. | 16:57 |
| *** djkonro has joined #openstack-glance | 16:58 | |
| nikhil | I am trying to think of it as tsymanczyk mentioned at the summit on granular ACLs | 16:58 |
| *** djkonro is now known as Guest18237 | 16:59 | |
| * rosmaita saw something shiny ... reading back through the discussion | 16:59 | |
| nikhil | based on what rosmaita says, I think that should work | 16:59 |
| sigmavirus24 | I have to run an errand, but I'll be back | 16:59 |
| nikhil | basically, visibility is like "change group" on your linux FS | 16:59 |
| nikhil | sigmavirus24: np, thanks for your feedback! | 16:59 |
| nikhil | so.. | 17:00 |
| nikhil | visibility == chgrp | 17:00 |
| nikhil | and | 17:00 |
| nikhil | members == chmod | 17:00 |
| sigmavirus24 | nikhil: it's never worked that way and I'm not sure I like that analogy | 17:00 |
| nikhil | ? | 17:00 |
| sigmavirus24 | members is more like chgrp | 17:00 |
| sigmavirus24 | changing owner more like chown | 17:01 |
| sigmavirus24 | visibility is more chmod | 17:01 |
| sigmavirus24 | i.e., file permissions determine what owner can do, group can do, everyoen else | 17:01 |
| sigmavirus24 | anyway, brb | 17:01 |
| nikhil | sigmavirus24: so here's more context | 17:01 |
| *** Guest18237 has quit IRC | 17:01 | |
| nikhil | sigmavirus24: ok | 17:01 |
| nikhil | rosmaita: if we go with sigmavirus24 then 'inherited' images would need to be a sub-part of membership (chgrp ie) | 17:02 |
| nikhil | rosmaita: not protected as that's already a thing in glance (protected flag to disallow unintended deletions) | 17:03 |
| nikhil | tsymanczyk: I realize that you've a meeting now | 17:05 |
| rosmaita | good point about 'protected' | 17:05 |
| nikhil | tsymanczyk: or may be you're already gone for it :) | 17:05 |
| tsymanczyk | thank you for the consideration, but it's been rescheduled for 4pm. so i am still here, just listening. | 17:05 |
| nikhil | :) | 17:05 |
| nikhil | actually, thinking a bit more | 17:07 |
| rosmaita | so could we work it to go with some kind of mask instead of keywords? | 17:07 |
| nikhil | the ch(own, grp, mod) is pure waste of analogy | 17:07 |
| nikhil | and that's my bad completely | 17:07 |
| nikhil | and I prolly mistook what tsymanczyk said at the summit | 17:08 |
| rosmaita | i think we shoudl explore it ... why bad? | 17:08 |
| nikhil | rosmaita: because, it's action based ie has pure CRUD on the ownership | 17:09 |
| nikhil | or membership etc | 17:09 |
| nikhil | rosmaita: for us, it's merely usage? | 17:09 |
| nikhil | us == sharing | 17:09 |
| nikhil | so... | 17:09 |
| *** jistr has quit IRC | 17:09 | |
| nikhil | I think we need to think of social newtworks when sharing is concerned | 17:10 |
| tsymanczyk | if a bitmask is a better idea then awesome, but i'm afraid i'm missing where the the four discrete states fall down (public, private, community, shared). is sigmavirus24's concern beyond the fact that "private" today (ie, not is_public) can still be shared if it has members? | 17:10 |
| tsymanczyk | perhaps i just don't have the background to see the problem. | 17:11 |
| *** mfedosin has quit IRC | 17:13 | |
| rosmaita | tsymanczyk: just read back through the discussion, i'm not sure either | 17:17 |
| rosmaita | my worry is whether 'inherited' images will work as another visibility, or whether they would cut across the 4 modes | 17:18 |
| rosmaita | nikhil: instead of 'inherited images' how about 'family images' ? | 17:19 |
| nikhil | rosmaita: am not hung up on the term :) | 17:20 |
| nikhil | rosmaita: though, I am trying to understand what image sharing means for us? | 17:20 |
| nikhil | and if inheritence is actually sharing | 17:20 |
| nikhil | if not, how do we ensure they do not cut across each other's domains | 17:21 |
| *** david-lyle_ is now known as david-lyle | 17:22 | |
| tsymanczyk | i'm not against adding inherited into the mix at all, i'd just like to ensure there's a real justification for adding complexity before doing so. | 17:22 |
| rosmaita | tsymanczyk: i agree, i just want to make sure we don't back ourselves into a corner | 17:23 |
| tsymanczyk | cool | 17:24 |
| nikhil | so, I think we need to process this in the backgroud a bit | 17:24 |
| nikhil | at the very least I need to | 17:24 |
| rosmaita | nikhil: i don't know if inheritance is actually sharing ... sharing seems intentional, inheritance seems automatic; although, there may be some things you don't want to expose to inheritance, so intention does show up there, too | 17:24 |
| nikhil | rosmaita: I dunno what those things are today though :) | 17:25 |
| nikhil | rosmaita: it is intentional in the sense that you do want to restrict the 'visibility' of the image to specific tenants | 17:27 |
| nikhil | rosmaita: but it's a bit more tightly couple with what's in keystone. vs. sharing is something glance dictates. | 17:28 |
| rosmaita | tsymanczyk: what i'm worried about is that we're building community images on top of shared images, and if our conceptual framework for shared images is flawed, we may make a big mess | 17:28 |
| rosmaita | nikhil: except if you want to *not* have a particular item be inherited, glance has to do that, not keystone | 17:30 |
| nikhil | rosmaita: agreed | 17:31 |
| tsymanczyk | totally agree. it's important that our fundamental design is solid. | 17:32 |
| rosmaita | tsymanczyk: i guess i'm worried because the original sharing model was based on enabling an "image marketplace" | 17:33 |
| rosmaita | so it's good for that, but maybe not other types of sharing | 17:33 |
| rosmaita | but community sharing is a pretty clear use case | 17:34 |
| rosmaita | we've complicated it a bit by adding the "bookmarking" feature, by hijacking current image sharing | 17:35 |
| nikhil | I need a whiteboard | 17:36 |
| nikhil | :) | 17:36 |
| *** dshakhray has joined #openstack-glance | 17:36 | |
| *** daemontool has joined #openstack-glance | 17:37 | |
| nikhil | rosmaita: tsymanczyk : I gtg but you guys can carry on. would like to catch up more next week and/or on the spec. | 17:37 |
| tsymanczyk | true. not to derail, but that also brings up another good point sigmavirus24 touched on. if we go ahead with the pending, rejected, accepted paradigm - how does teh consumer "know" they've been invited? and are we sure that we need these state transitions? (an alternate suggestion, off the cuff) what if for a shared image, the image owner can add members. | 17:38 |
| tsymanczyk | but for a community image, it must be the consumer to add themselves as a member. | 17:38 |
| tsymanczyk | nikhil : sounds good. i'll be monitoring the spec and will be happy to back and forth there, and will certainly be back here next week same time. | 17:38 |
| rosmaita | tsymanczyk: for community images, the consumer does add themself | 17:39 |
| *** daemontool_ has quit IRC | 17:39 | |
| tsymanczyk | true. i meant to say that ONLY the consumer can add themselves in the community case (not the provider). then we don't need the concept of invited. because either the consumer wants it and adds themselves, or they don't and they don't. | 17:40 |
| sigmavirus24 | So I think IRC is failing us for this discussion | 17:44 |
| *** e0ne has joined #openstack-glance | 17:46 | |
| sigmavirus24 | Here's my thing, right now we not publicly shared images that are shared are private. Going forward, they will need to be changed to "shared" before they can be shared privately? | 17:46 |
| sigmavirus24 | I really don't like that | 17:46 |
| tsymanczyk | it's the explicit two-step that you don't like? | 17:46 |
| sigmavirus24 | I'm trying to understand the needs to have a new visibility state when I feel like it's unnecessary and you can get the same benefits without having users add themselves to otherwise hidden images | 17:46 |
| sigmavirus24 | tsymanczyk: that's part of it | 17:46 |
| sigmavirus24 | Also, I haven't read teh entire spec, but members are still going to be project/tenant ids, right? So if a user in a project hides the image, does that hide it for everyone else in their project? | 17:47 |
| tsymanczyk | in what case in the proposal does a user add themselves to an otherwise hidden image? | 17:47 |
| sigmavirus24 | I think because of how members work, that answer is yes | 17:47 |
| sigmavirus24 | tsymanczyk: sorry, confusing the community visibility discussion above of bookmarking with the misunderstanding of member_status from before | 17:48 |
| sigmavirus24 | also member_status as a name for whether a member has chosen to see an image in their image list is very confusing (as well as the language used to describe that state change) | 17:49 |
| sigmavirus24 | I think I just need to set time aside later to read the entirety of the spec though | 17:49 |
| sigmavirus24 | What i've been asked to read seems overly complicated | 17:49 |
| sigmavirus24 | Which already describes enough of glance | 17:49 |
| sigmavirus24 | And an image marketplace sounds to me like murano | 17:49 |
| sigmavirus24 | Although images are only one part of the app catalog/murano | 17:50 |
| tsymanczyk | that point regarding member_status i think you've convinced me needs some thought at the very least. unnecessary complexity is bad let's avoid it. | 17:50 |
| sigmavirus24 | I'm sorry, I don't mean to be a negative person. I wasn't at the summit so I missed a lot of the context that y'all have | 17:50 |
| sigmavirus24 | But coming at this with a fresh set of eyes, I see potentially unnecessary backwards incompatibility | 17:50 |
| sigmavirus24 | And confusing complexity | 17:51 |
| sigmavirus24 | And perhaps we need a more nuanced idea of visibility | 17:51 |
| tsymanczyk | i really appreciate your input. i'm not taking it negatively. | 17:51 |
| sigmavirus24 | Because as it is described now, it seems like what we want to describe is "public:not-community", "public:community", "private:no-members", "private:has-members" | 17:52 |
| sigmavirus24 | So it feels to me right this second that we're describing specializations of the existing visibilities which is cool | 17:52 |
| sigmavirus24 | nuance is good, but it's a bit confusing right now and I really need time to digest it :) | 17:52 |
| sigmavirus24 | That said, I think we also have a good amount of time until specs need to be approved so I don't want you to feel rushed | 17:53 |
| *** ducttape_ has quit IRC | 17:53 | |
| tsymanczyk | i think we have the same ideas and goals, with different terminology. "public:not-community" -> public. "public:community" -> community. "private:no-members" -> private. "private:has-members" -> shared. | 17:53 |
| tsymanczyk | i don't see how they're different. | 17:54 |
| tsymanczyk | but i understand what you're saying, you want additional time to read and think. so i'm not trying to push the discussion RIGHT NOW. but i also am having trouble seeing where the disagreement is. | 17:56 |
| tsymanczyk | getting the fundamentals right is very important, and yes you're right we have a lot of time left to make sure we do that. | 17:56 |
| *** mtanino has quit IRC | 17:57 | |
| openstackgerrit | Hemanth Makkapati proposed openstack/glance: Improve help text of registry client context opts https://review.openstack.org/319396 | 18:02 |
| rosmaita | tsymanczyk: sigmavirus24: sorry, was away for a bit | 18:07 |
| rosmaita | tsymanczyk: i think you are right, there is more agreement bet you and sigmavirus24 than ian seems to think | 18:08 |
| rosmaita | sigmavirus24: it would be great if you could read through the spec | 18:08 |
| rosmaita | i am also worried about over-complicating things | 18:08 |
| sigmavirus24 | rosmaita: so I'm very concerned about how new visibilities will affect people's scripts though | 18:08 |
| sigmavirus24 | specifically shared might break things | 18:08 |
| rosmaita | well, it's kind of funny | 18:09 |
| rosmaita | a lot of people think there's already a 'shared' visibility | 18:09 |
| sigmavirus24 | My point with the specialization is just that, I feel like "shared" is a kind of private | 18:09 |
| rosmaita | well, in the ownership sense you are correct | 18:10 |
| rosmaita | part of what we want to do is separate vizibility from ownership | 18:10 |
| rosmaita | sigmavirus24: you might want to take a quick look at this: https://developer.rackspace.com/docs/cloud-images/v2/developer-guide/#image-sharing | 18:12 |
| rosmaita | there used to be something similar upstream but i don't know where it went | 18:12 |
| rosmaita | it outlines the ground rules for current glance v2 image sharing | 18:12 |
| *** mtanino has joined #openstack-glance | 18:14 | |
| *** ducttape_ has joined #openstack-glance | 18:16 | |
| *** pushkaru has quit IRC | 18:18 | |
| tsymanczyk | is there a use case where we'd WANT for a consumer to have "pending" membership status on a community image (or any other state beyond accepted)? been wracking my brain, haven't come up with one. | 18:22 |
| *** julim has quit IRC | 18:27 | |
| *** sgotliv_ has quit IRC | 18:31 | |
| rosmaita | tsymanczyk: i don't think so | 18:34 |
| rosmaita | tsymanczyk: i mean, i can't think of a use case, either | 18:35 |
| *** groen692 has quit IRC | 18:35 | |
| sigmavirus24 | rosmaita: I'm guessing I've used v2 so little that I didn't know member_status wasa lready a thing? | 18:40 |
| sigmavirus24 | (sorry I'm juggling like 4 things right now) | 18:42 |
| rosmaita | it has an ancient and honorable history | 18:42 |
| sigmavirus24 | and learning things about SSH I didn't want to learn | 18:42 |
| rosmaita | i think it's what made 2.0 -> 2.1 | 18:42 |
| sigmavirus24 | I think 2.0 and 2.1 are long before my tenure on the project, let alone Rackspace :P | 18:42 |
| *** sdake has joined #openstack-glance | 18:46 | |
| *** ihrachys has quit IRC | 18:47 | |
| *** pdeore has quit IRC | 18:49 | |
| *** dmk0202 has joined #openstack-glance | 18:50 | |
| *** ninag_ has quit IRC | 19:02 | |
| *** ninag has joined #openstack-glance | 19:02 | |
| openstackgerrit | Hemanth Makkapati proposed openstack/glance: Improve help text of registry server opts https://review.openstack.org/318910 | 19:06 |
| *** ninag has quit IRC | 19:06 | |
| *** pushkaru has joined #openstack-glance | 19:06 | |
| *** julim has joined #openstack-glance | 19:20 | |
| *** ayoung has quit IRC | 19:29 | |
| *** ninag has joined #openstack-glance | 19:35 | |
| *** dshakhray has quit IRC | 19:45 | |
| *** btully has quit IRC | 19:46 | |
| *** ihrachys has joined #openstack-glance | 20:04 | |
| *** e0ne has quit IRC | 20:08 | |
| *** e0ne has joined #openstack-glance | 20:12 | |
| *** btully has joined #openstack-glance | 20:12 | |
| *** btully has quit IRC | 20:17 | |
| *** ninag has quit IRC | 20:19 | |
| *** TravT_ has joined #openstack-glance | 20:21 | |
| *** TravT has quit IRC | 20:22 | |
| *** sdake has quit IRC | 20:22 | |
| *** sdake has joined #openstack-glance | 20:22 | |
| *** TravT_ has quit IRC | 20:24 | |
| *** TravT has joined #openstack-glance | 20:25 | |
| *** sdake has quit IRC | 20:25 | |
| *** ayoung has joined #openstack-glance | 20:26 | |
| *** smatzek has quit IRC | 20:26 | |
| *** e0ne has quit IRC | 20:32 | |
| *** ninag has joined #openstack-glance | 20:32 | |
| *** delattec has quit IRC | 20:33 | |
| *** djkonro has joined #openstack-glance | 20:39 | |
| *** pushkaru has quit IRC | 20:41 | |
| *** sdake has joined #openstack-glance | 20:45 | |
| *** djkonro has quit IRC | 20:47 | |
| *** pushkaru has joined #openstack-glance | 20:50 | |
| *** raildo is now known as raildo-afk | 20:53 | |
| *** btully has joined #openstack-glance | 21:02 | |
| *** smatzek has joined #openstack-glance | 21:05 | |
| *** ninag has quit IRC | 21:06 | |
| *** btully has quit IRC | 21:07 | |
| *** btully has joined #openstack-glance | 21:09 | |
| *** djkonro has joined #openstack-glance | 21:19 | |
| *** ninag has joined #openstack-glance | 21:19 | |
| *** djkonro is now known as Guest31597 | 21:20 | |
| *** rcernin has joined #openstack-glance | 21:20 | |
| *** daemontool has quit IRC | 21:22 | |
| *** smatzek has quit IRC | 21:26 | |
| *** Guest31597 has quit IRC | 21:26 | |
| *** ametts has quit IRC | 21:31 | |
| nikhil | jokke_: rosmaita : I guess I missed the ping from morning on etherpad discussion and realized now that I need to respond. We'd chat monday as some things strike important there. | 21:33 |
| *** pushkaru has quit IRC | 21:35 | |
| *** ducttape_ has quit IRC | 21:36 | |
| *** Guest31597 has joined #openstack-glance | 21:38 | |
| *** Guest31597 has quit IRC | 21:46 | |
| openstackgerrit | Hemanth Makkapati proposed openstack/glance: Improve help text of quota opts https://review.openstack.org/319467 | 21:47 |
| *** dmk0202 has quit IRC | 21:57 | |
| *** Guest31597 has joined #openstack-glance | 21:58 | |
| *** sdake has quit IRC | 21:59 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 22:01 | |
| *** burgerk_ has quit IRC | 22:04 | |
| *** Guest31597 has quit IRC | 22:08 | |
| *** sdake has joined #openstack-glance | 22:09 | |
| *** Guest31597 has joined #openstack-glance | 22:23 | |
| *** stevemar is now known as notstevemar | 22:24 | |
| *** ninag has quit IRC | 22:26 | |
| *** Guest31597 has quit IRC | 22:34 | |
| *** ninag has joined #openstack-glance | 22:37 | |
| *** jamielennox|away is now known as jamielennox | 22:48 | |
| *** krotscheck is now known as krotscheck_dcm | 22:57 | |
| *** Guest31597 has joined #openstack-glance | 23:01 | |
| *** Guest31597 has quit IRC | 23:05 | |
| *** marcusvrn_ has quit IRC | 23:13 | |
| *** ihrachys has quit IRC | 23:19 | |
| *** timburke has quit IRC | 23:27 | |
| *** timburke has joined #openstack-glance | 23:38 | |
| *** turvey has joined #openstack-glance | 23:42 | |
| *** turvey has left #openstack-glance | 23:48 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!