| *** takedakn1 has joined #openstack-glance | 00:03 | |
| *** takedakn has quit IRC | 00:07 | |
| *** takedakn1 is now known as takedakn | 00:07 | |
| *** takedakn1 has joined #openstack-glance | 00:10 | |
| *** takedakn has quit IRC | 00:11 | |
| *** takedakn1 is now known as takedakn | 00:11 | |
| *** ducttape_ has joined #openstack-glance | 00:16 | |
| *** ducttape_ has quit IRC | 00:18 | |
| *** goodygum has quit IRC | 00:19 | |
| openstackgerrit | Merged openstack/glance_store: Release notes 0.9.0 and corrected library version https://review.openstack.org/219886 | 00:20 |
|---|---|---|
| *** mtanino_ has quit IRC | 00:21 | |
| *** goodygum has joined #openstack-glance | 00:21 | |
| *** achanda has joined #openstack-glance | 00:28 | |
| *** takedakn1 has joined #openstack-glance | 00:40 | |
| *** takedakn has quit IRC | 00:44 | |
| *** takedakn1 has quit IRC | 00:46 | |
| *** achanda has quit IRC | 00:51 | |
| *** achanda has joined #openstack-glance | 00:52 | |
| *** achanda has quit IRC | 00:52 | |
| *** julim has joined #openstack-glance | 01:05 | |
| *** zul has quit IRC | 01:14 | |
| *** mingdang1 has joined #openstack-glance | 01:14 | |
| *** kebray has joined #openstack-glance | 01:15 | |
| *** julim has quit IRC | 01:15 | |
| *** zul has joined #openstack-glance | 01:17 | |
| *** julim has joined #openstack-glance | 01:20 | |
| *** skylerberg has quit IRC | 01:22 | |
| *** ducttape_ has joined #openstack-glance | 01:29 | |
| *** ducttape_ has quit IRC | 01:30 | |
| *** ducttape_ has joined #openstack-glance | 01:31 | |
| *** smatzek has joined #openstack-glance | 01:33 | |
| *** sdake has quit IRC | 01:36 | |
| *** achanda has joined #openstack-glance | 01:44 | |
| *** julim has quit IRC | 01:47 | |
| *** boris-42 has quit IRC | 01:50 | |
| openstackgerrit | Teruaki Ishizaki proposed openstack/glance_store: Sheepdog: improve operations for storing glance-image https://review.openstack.org/205942 | 02:06 |
| *** TravT has joined #openstack-glance | 02:08 | |
| *** ducttape_ has quit IRC | 02:15 | |
| *** smatzek has quit IRC | 02:17 | |
| *** julim has joined #openstack-glance | 02:26 | |
| *** sdake has joined #openstack-glance | 02:28 | |
| *** Kennan has quit IRC | 02:29 | |
| *** Kennan has joined #openstack-glance | 02:30 | |
| *** sdake_ has joined #openstack-glance | 02:31 | |
| *** sdake has quit IRC | 02:34 | |
| *** haomaiwang has quit IRC | 02:36 | |
| *** haomaiwang has joined #openstack-glance | 02:38 | |
| *** haomaiwang has quit IRC | 02:57 | |
| *** haomaiwa_ has joined #openstack-glance | 03:00 | |
| *** bkopilov has quit IRC | 03:05 | |
| *** harshs has joined #openstack-glance | 03:07 | |
| *** eglute has quit IRC | 03:27 | |
| *** achanda has quit IRC | 03:28 | |
| *** erhudy1 has quit IRC | 03:29 | |
| *** sigmavirus24 has quit IRC | 03:30 | |
| *** d34dh0r53 has quit IRC | 03:30 | |
| *** jroll has quit IRC | 03:30 | |
| *** sabari has quit IRC | 03:30 | |
| *** d34dh0r53 has joined #openstack-glance | 03:31 | |
| *** sabari has joined #openstack-glance | 03:31 | |
| *** eglute has joined #openstack-glance | 03:32 | |
| *** sigmavirus24 has joined #openstack-glance | 03:33 | |
| *** jroll has joined #openstack-glance | 03:33 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 03:34 | |
| *** tpeoples has joined #openstack-glance | 03:35 | |
| *** dims has joined #openstack-glance | 03:35 | |
| *** sdake_ is now known as sdake | 03:42 | |
| *** spzala has quit IRC | 03:43 | |
| *** tpeoples has quit IRC | 03:46 | |
| *** jecarey has quit IRC | 03:46 | |
| *** julim has quit IRC | 03:54 | |
| *** boris-42 has joined #openstack-glance | 03:59 | |
| *** dims has quit IRC | 04:00 | |
| *** links has joined #openstack-glance | 04:02 | |
| *** rajesht has joined #openstack-glance | 04:03 | |
| *** kebray has quit IRC | 04:11 | |
| *** baojg has joined #openstack-glance | 04:11 | |
| *** kebray has joined #openstack-glance | 04:12 | |
| *** baojg has quit IRC | 04:31 | |
| *** tpeoples has joined #openstack-glance | 04:37 | |
| *** ayoung has quit IRC | 04:44 | |
| *** tpeoples has quit IRC | 04:55 | |
| *** bkopilov has joined #openstack-glance | 04:57 | |
| *** dims has joined #openstack-glance | 04:59 | |
| *** dims has quit IRC | 05:06 | |
| *** sdake_ has joined #openstack-glance | 05:13 | |
| *** sdake has quit IRC | 05:17 | |
| *** Kennan2 has joined #openstack-glance | 05:24 | |
| *** Kennan has quit IRC | 05:25 | |
| openstackgerrit | Wayne Okuma proposed openstack/glance: Glance metadef tables need unique constraints. https://review.openstack.org/195820 | 05:32 |
| *** achanda has joined #openstack-glance | 05:34 | |
| *** achanda has quit IRC | 05:40 | |
| *** ishant has joined #openstack-glance | 05:43 | |
| *** sdake has joined #openstack-glance | 05:48 | |
| *** sdake_ has quit IRC | 05:52 | |
| *** sdake_ has joined #openstack-glance | 05:54 | |
| *** sdake has quit IRC | 05:57 | |
| *** groen692 has joined #openstack-glance | 05:58 | |
| *** achanda has joined #openstack-glance | 06:00 | |
| *** achanda has quit IRC | 06:19 | |
| *** harshs has quit IRC | 06:29 | |
| *** sgotliv has joined #openstack-glance | 06:30 | |
| openstackgerrit | Flavio Percoco proposed openstack/python-glanceclient: Consider `--os-token` when using v2 https://review.openstack.org/219686 | 06:34 |
| openstackgerrit | Flavio Percoco proposed openstack/python-glanceclient: Check if v2 is available and fallback https://review.openstack.org/219469 | 06:34 |
| *** abhishekk has joined #openstack-glance | 06:43 | |
| *** achanda has joined #openstack-glance | 06:48 | |
| *** peristeri has joined #openstack-glance | 06:58 | |
| *** dedery has joined #openstack-glance | 07:26 | |
| openstackgerrit | Flavio Percoco proposed openstack/python-glanceclient: Consider `--os-token` when using v2 https://review.openstack.org/219686 | 07:58 |
| *** dedery has quit IRC | 08:02 | |
| *** links has quit IRC | 08:02 | |
| *** kebray has quit IRC | 08:22 | |
| *** dedery_ has joined #openstack-glance | 08:26 | |
| *** achanda has quit IRC | 08:36 | |
| *** jistr has joined #openstack-glance | 08:37 | |
| *** dedery_ has quit IRC | 08:55 | |
| *** mingdang1 has quit IRC | 08:57 | |
| *** d0ugal has left #openstack-glance | 09:02 | |
| *** ccard has joined #openstack-glance | 09:08 | |
| *** MarcF has quit IRC | 09:19 | |
| *** belmoreira has joined #openstack-glance | 09:26 | |
| *** MarcF has joined #openstack-glance | 09:31 | |
| *** e0ne has joined #openstack-glance | 09:32 | |
| openstackgerrit | Niall Bunting proposed openstack/python-glanceclient: image-list backwards compatibility for v2 https://review.openstack.org/219802 | 09:37 |
| openstackgerrit | Takashi NATSUME proposed openstack/python-glanceclient: Add a condition to avoid mistaking the version string https://review.openstack.org/218126 | 09:42 |
| *** e0ne has quit IRC | 10:04 | |
| *** MarcF has quit IRC | 10:04 | |
| *** e0ne has joined #openstack-glance | 10:08 | |
| *** MarcF has joined #openstack-glance | 10:16 | |
| openstackgerrit | Niall Bunting proposed openstack/python-glanceclient: image-list backwards compatibility for v2 https://review.openstack.org/219802 | 10:17 |
| *** MattMan has quit IRC | 10:23 | |
| *** MattMan has joined #openstack-glance | 10:24 | |
| *** links has joined #openstack-glance | 10:31 | |
| openstackgerrit | Niall Bunting proposed openstack/python-glanceclient: image-list backwards compatibility for v2 https://review.openstack.org/219802 | 10:37 |
| openstackgerrit | Kairat Kushaev proposed openstack/glance: Check value format when uploading blob value https://review.openstack.org/219262 | 10:37 |
| *** smatzek has joined #openstack-glance | 10:40 | |
| *** Kennan2 is now known as Kennan_on_vacati | 10:44 | |
| *** Kennan_on_vacati is now known as Kennan_Vacation | 10:45 | |
| *** aix has quit IRC | 10:46 | |
| openstackgerrit | Niall Bunting proposed openstack/python-glanceclient: image-list backwards compatibility for v2 https://review.openstack.org/219802 | 10:52 |
| *** aix has joined #openstack-glance | 11:15 | |
| *** ishant has quit IRC | 11:19 | |
| *** takedakn has joined #openstack-glance | 11:25 | |
| *** sayali has quit IRC | 11:26 | |
| *** MarcF has quit IRC | 11:29 | |
| openstackgerrit | Flavio Percoco proposed openstack/python-glanceclient: Consider `--os-token` when using v2 https://review.openstack.org/219686 | 11:35 |
| *** takedakn has quit IRC | 11:36 | |
| *** takedakn has joined #openstack-glance | 11:36 | |
| *** pbourke has quit IRC | 11:36 | |
| *** pbourke has joined #openstack-glance | 11:37 | |
| *** julim has joined #openstack-glance | 11:37 | |
| openstackgerrit | Merged openstack/glance: Add image signing verification https://review.openstack.org/183137 | 11:39 |
| *** takedakn1 has joined #openstack-glance | 11:41 | |
| *** MarcF has joined #openstack-glance | 11:43 | |
| *** takedakn has quit IRC | 11:44 | |
| *** takedakn1 is now known as takedakn | 11:44 | |
| *** bkopilov has quit IRC | 11:51 | |
| *** takedakn has quit IRC | 11:52 | |
| openstackgerrit | Flavio Percoco proposed openstack/python-glanceclient: Don't make `help` require auth parameters https://review.openstack.org/220096 | 11:54 |
| *** ducttape_ has joined #openstack-glance | 12:10 | |
| *** bfic has joined #openstack-glance | 12:11 | |
| openstackgerrit | Takashi NATSUME proposed openstack/python-glanceclient: Add a condition to avoid mistaking the version string https://review.openstack.org/218126 | 12:19 |
| *** spzala has joined #openstack-glance | 12:19 | |
| openstackgerrit | Niall Bunting proposed openstack/python-glanceclient: Move metadef into sub-commands https://review.openstack.org/218864 | 12:23 |
| openstackgerrit | Takashi NATSUME proposed openstack/python-glanceclient: Add a condition to avoid mistaking the version string https://review.openstack.org/218126 | 12:27 |
| *** ducttape_ has quit IRC | 12:31 | |
| openstackgerrit | Kairat Kushaev proposed openstack/glance: Check value format when uploading blob value https://review.openstack.org/219262 | 12:33 |
| *** burgerk has joined #openstack-glance | 12:33 | |
| *** dims has joined #openstack-glance | 12:41 | |
| *** sayali has joined #openstack-glance | 12:42 | |
| *** jw92 has joined #openstack-glance | 12:45 | |
| *** jw92 has left #openstack-glance | 12:52 | |
| *** sayali has quit IRC | 12:52 | |
| *** chutwig has joined #openstack-glance | 12:55 | |
| *** sayali has joined #openstack-glance | 12:56 | |
| openstackgerrit | Niall Bunting proposed openstack/python-glanceclient: image-list backwards compatibility for v2 https://review.openstack.org/219802 | 12:56 |
| *** qwebirc45097 has joined #openstack-glance | 12:59 | |
| *** jw92_ has joined #openstack-glance | 13:00 | |
| *** qwebirc45097 has quit IRC | 13:00 | |
| openstackgerrit | Stuart McLaren proposed openstack/python-glanceclient: Add a condition to avoid mistaking the version string https://review.openstack.org/218126 | 13:09 |
| *** edmondsw has joined #openstack-glance | 13:13 | |
| *** sayali has quit IRC | 13:16 | |
| *** sayali has joined #openstack-glance | 13:22 | |
| *** sayali has quit IRC | 13:24 | |
| *** mingdang1 has joined #openstack-glance | 13:27 | |
| *** sayali has joined #openstack-glance | 13:27 | |
| *** groen692 has quit IRC | 13:32 | |
| *** groen692 has joined #openstack-glance | 13:32 | |
| *** jecarey has joined #openstack-glance | 13:34 | |
| dhellmann | glance-store 0.9.0 is tagged | 13:39 |
| *** links has quit IRC | 13:40 | |
| *** burgerk has quit IRC | 13:41 | |
| *** bfic has left #openstack-glance | 13:47 | |
| jokke_ | thanks dhellmann! | 13:52 |
| *** peristeri has quit IRC | 13:53 | |
| *** bpoulos has joined #openstack-glance | 13:53 | |
| *** jroll has quit IRC | 13:54 | |
| *** jroll has joined #openstack-glance | 13:54 | |
| *** sayali has quit IRC | 13:57 | |
| *** tpeoples has joined #openstack-glance | 13:58 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 13:59 | |
| *** sayali has joined #openstack-glance | 13:59 | |
| nikhil_k | Courtesy meeting reminder: ativelkov, cpallares, esheffield, flaper87, flwang1, hemanthm, ivasilevskaya, jokke_, kragniz, lakshmiS, mclaren, mfedosin, nikhil_k, Nikolay_St, Olena, pennerc, rosmaita, sigmavirus24, sabari, TravT, zhiyan, pkoniszewski, krykowski, ajayaa, GB21, bpoulos, harshs, abhishekk, bunting, dshakhray | 14:00 |
| *** MarcF has quit IRC | 14:00 | |
| *** ducttape_ has joined #openstack-glance | 14:02 | |
| *** MarcF has joined #openstack-glance | 14:03 | |
| *** burgerk has joined #openstack-glance | 14:03 | |
| *** harshs has joined #openstack-glance | 14:10 | |
| *** vijendar has joined #openstack-glance | 14:17 | |
| *** vijendar has quit IRC | 14:17 | |
| *** vijendar has joined #openstack-glance | 14:17 | |
| *** MarcF has quit IRC | 14:19 | |
| *** ayoung has joined #openstack-glance | 14:24 | |
| *** spzala has quit IRC | 14:28 | |
| openstackgerrit | Brian Rosmaita proposed openstack/glance: Add tasks info to glance documentation https://review.openstack.org/220166 | 14:30 |
| *** MarcF has joined #openstack-glance | 14:30 | |
| *** bkopilov has joined #openstack-glance | 14:32 | |
| *** devlaps has quit IRC | 14:33 | |
| *** mingdang1 has quit IRC | 14:34 | |
| *** mtanino has joined #openstack-glance | 14:35 | |
| *** chutwig has quit IRC | 14:36 | |
| *** xylan_kong has left #openstack-glance | 14:39 | |
| *** sayali has quit IRC | 14:40 | |
| *** TravT_ has joined #openstack-glance | 14:49 | |
| *** spzala has joined #openstack-glance | 14:52 | |
| *** TravT has quit IRC | 14:52 | |
| *** wznoinsk has quit IRC | 14:52 | |
| *** mtanino has quit IRC | 14:53 | |
| h0mer | what is the best way to export images out of glance and onto my local machine? | 14:55 |
| *** kebray has joined #openstack-glance | 15:00 | |
| *** abhishekk has quit IRC | 15:01 | |
| nikhil_k | o/ | 15:01 |
| *** TravT_ is now known as TravT | 15:01 | |
| jokke_ | h0mer: image-download | 15:02 |
| *** jistr is now known as jistr|call | 15:02 | |
| bunting | nikhil_k: My patch quietly deals with these old values | 15:02 |
| bunting | which if you expect a script to run, would you not rather have that behaviour than it just break? | 15:03 |
| *** mclaren has joined #openstack-glance | 15:03 | |
| *** akanksha has joined #openstack-glance | 15:04 | |
| nikhil_k | bunting: so my question is: why should we update the client to make it compact with the older API just to unbrak scripts. The version bump loses meaning!? | 15:04 |
| nikhil_k | It's matter of packaging that a specific version is installed and used | 15:05 |
| *** sayali has joined #openstack-glance | 15:05 | |
| nikhil_k | jokke_: your https://review.openstack.org/#/c/210025/ is being the only wait for L-3 | 15:05 |
| nikhil_k | jokke_: or do you not care if this goes in L3 or not? | 15:06 |
| nikhil_k | flaper87: ^ | 15:06 |
| bunting | I would take the version bump as proper way to do it, but we should support older scripts, at least with a warning before they just break | 15:06 |
| bunting | The version bump would not lose meaning, as it marks a shift, but gives people time to move to the new set | 15:07 |
| nikhil_k | bunting: ok, how should we handle the warning!? I think we may have a million users for clients who need to update their scripts? | 15:07 |
| nikhil_k | Just trying to understand | 15:08 |
| nikhil_k | Basically how long of a communication gap should we propose? | 15:08 |
| nikhil_k | and what needs to be the overlap? | 15:08 |
| nikhil_k | or if there's way we can release 0.20.0 that would solve the issue? | 15:08 |
| nikhil_k | Don't want to be in catch22 when people say 1.0.0 has all those commands and now we can never revert them | 15:09 |
| bunting | Yeah, we should at least have an overlap the duration is an intresting question | 15:09 |
| *** jw92_ has quit IRC | 15:09 | |
| bunting | Well the old commands should be hidden in my opinion, so if you do glance help they would not show up, but still work | 15:10 |
| *** sayali has quit IRC | 15:10 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 15:12 | |
| bunting | as that means you would slowly update your scripts as if you make changes, and use the new help, it will basicly force you to update as only the new commands are shown | 15:12 |
| *** sigmavirus24_awa is now known as sigmavirus24 | 15:13 | |
| *** MarcF has quit IRC | 15:13 | |
| *** chutwig has joined #openstack-glance | 15:14 | |
| *** harshs has quit IRC | 15:17 | |
| nikhil_k | bunting: I see | 15:19 |
| jokke_ | nikhil_k: that and the needed by is one and the same if it goes in to l3 or RC1 ... would be just nice to get that fixed before release | 15:21 |
| *** annegentle has joined #openstack-glance | 15:21 | |
| nikhil_k | jokke_: it _may_ go in rc1 if doc team doesn't come screaming | 15:22 |
| jokke_ | nikhil_k: in that case we need it to l3 | 15:22 |
| jokke_ | that is sec vulnerability waiting to be plugged | 15:22 |
| nikhil_k | mclaren: bunting : So, the above approach makes sense. I think if we can ensure that out overlap period is communicated well and we say that 1.x.x series will support that overlap.. | 15:22 |
| *** sayali has joined #openstack-glance | 15:22 | |
| nikhil_k | then we can define the 2.0.0 release and remove the older ones | 15:23 |
| nikhil_k | jokke_: flaper87 ^ | 15:23 |
| nikhil_k | jokke_: yep | 15:23 |
| jokke_ | nikhil_k: so that change and the one depending on that | 15:23 |
| nikhil_k | jokke_: once that merges, we need 2 +2s on the dependent | 15:23 |
| bunting | nikhil_k: Yeah, and 2.x.x would need to support 1.x.x to give people time to change | 15:24 |
| jokke_ | nikhil_k, bunting, mclaren: one reminder ... any of your "backwards compatibility" changes cannot then affect if --os-image-api-version 2 or env[OS_IMAGE_API_VERSION]=2 is specified | 15:24 |
| jokke_ | as that's just breaking backwards comaptibility for all those who are relying on v2 api and it's behavior | 15:25 |
| *** kebray has joined #openstack-glance | 15:25 | |
| *** dims has quit IRC | 15:26 | |
| nikhil_k | bunting: I don't think we can allow those rolled over from 0.x.x to 1.x.x into 2.x.x but sure those that are new in 1.x.x | 15:26 |
| nikhil_k | jokke_: that's a good point | 15:26 |
| bunting | bunting: I meant the new 1.x.x things, so there is always like one overlap | 15:26 |
| nikhil_k | bunting: ^ | 15:26 |
| nikhil_k | hidden commands work to save the script breakage | 15:27 |
| bunting | We tend to only add in the old commands, what would be the harm of supporting it when it api 2 is specified? | 15:27 |
| nikhil_k | but not the specified API behavior is dented auto-magically | 15:27 |
| nikhil_k | because, it's confusing | 15:28 |
| nikhil_k | anyone who does glance --debug my_command is bound to get confused | 15:28 |
| jokke_ | bunting: I'm pretty sure there is 3-4 output "bugs" on that same tag for "backwards compatibility" | 15:28 |
| *** MarcF has joined #openstack-glance | 15:28 | |
| bunting | i understand output, however I dont think we can make that work as easily to be backwards compatable, because then we would have to change the v2 behaviour | 15:29 |
| bunting | but in the case off adding commands, im not sure it would be too much a problem | 15:30 |
| nikhil_k | as long as we hide them at specified api version | 15:31 |
| jokke_ | and every single one of those needs to print the warning out that it's using non-supported command for the API and will not work in future | 15:31 |
| jokke_ | to stderr | 15:32 |
| jokke_ | we do not want to get stuck with them | 15:32 |
| nikhil_k | I don't think we promise that | 15:32 |
| *** sayali has quit IRC | 15:32 | |
| bunting | We can hide them if its specified | 15:33 |
| jokke_ | nikhil_k: ? | 15:33 |
| nikhil_k | jokke_: promise to support them forever | 15:33 |
| jokke_ | nikhil_k: yeah, that's my point with those alerts | 15:33 |
| nikhil_k | we don't make such promise** | 15:33 |
| nikhil_k | yeah, I don't know how much of those scripted users will rely on stderr but sure why not if it helps!! | 15:34 |
| bunting | Yeah, i think we should warn on each one used, as that will push operators to use the new version | 15:34 |
| nikhil_k | the things with scripts is that they are bound to break with upgrades | 15:35 |
| jokke_ | nikhil_k: so if we do not want to stick with them we need to inform users about that ;) | 15:35 |
| nikhil_k | so making things explicit should be our approach | 15:35 |
| jokke_ | nikhil_k: and if they are hidden that cannot be the help text :P | 15:35 |
| *** dims has joined #openstack-glance | 15:36 | |
| *** sayali has joined #openstack-glance | 15:36 | |
| *** mtanino has joined #openstack-glance | 15:36 | |
| bunting | They do break, but we should make sure that is infrequent as we can | 15:36 |
| nikhil_k | jokke_: I don't know how many operators rely on the WARNINGs spit out everytime script if run. It really depends on how advanced your CI/CD is | 15:36 |
| nikhil_k | bunting: we can't guarantee that. | 15:37 |
| jokke_ | nikhil_k: but then they at least can't claim that we did not tell them | 15:38 |
| nikhil_k | I think we are being nice by adding a hidden command, so much like a deprecation period for them. | 15:38 |
| bunting | We can't guarantee that it won't break, but we can certainly minimise the damage | 15:38 |
| nikhil_k | The best way to avoid breakage is we give explicit info on how/when to upgrade. | 15:38 |
| jokke_ | nikhil_k: and what changes between versions | 15:39 |
| nikhil_k | It should be very similar to a wiki page that says something like: our default version is going to be v2, we will support v1 and have reduced support for v1 for so and so. On this date, we stop v1 support completely (by default). | 15:39 |
| nikhil_k | and in case of openstack, we can say 2.0.0 is when we stop and the tentative release period is date_range | 15:40 |
| jokke_ | nikhil_k: I think we need to support v1 in client bit longer than in the server side | 15:41 |
| mclaren | jokke_: +1 | 15:43 |
| nikhil_k | jokke_: sorry, I did not get it! if there's no v1 on server why do we need to support v1 on client? | 15:43 |
| mclaren | the person running the server is not the same persone installing the client code | 15:43 |
| jokke_ | nikhil_k: the client needs to support supported servers, no? | 15:44 |
| jokke_ | nikhil_k: so our grenade should blow up right away if we remove v1 support same time from client and server | 15:44 |
| *** jistr|call is now known as jistr | 15:44 | |
| *** mclaren has quit IRC | 15:45 | |
| flaper87 | here now, folks! Sorry I had another meeting | 15:47 |
| flaper87 | anything I can help with? | 15:47 |
| nikhil_k | jokke_: I see. So, if I understand it correctly - you mean to say that there should be a deprecation period | 15:47 |
| nikhil_k | flaper87: scroll back to see the proposed scheme for back compact commands | 15:48 |
| jokke_ | nikhil_k: I think if we remove api support from client we cannot do it without deprecation period | 15:48 |
| bunting | If we change the default version, i think we need a depreciation period as well | 15:49 |
| nikhil_k | jokke_: because, if we remove v1 from server and keep client v1 for a long time (say 12 months), it doesn't make much sense to me. | 15:49 |
| jokke_ | nikhil_k: and thus we should add that deprecation warning for any backwards comaptibility work we do here because otherwise they are considered as de facto behavior for that api | 15:49 |
| jokke_ | nikhil_k: I think it's 1 version what grenade tests | 15:50 |
| nikhil_k | sure, but upgrading grenade tests would need to be done around the same time as client release | 15:51 |
| bunting | nikhil_k: I think servers will upgrade more slowly, than clients so you need that 12 month support period | 15:52 |
| nikhil_k | by that logic we need smaller support period | 15:52 |
| nikhil_k | (for clients) | 15:53 |
| bunting | why would that be? | 15:54 |
| * flaper87 reads the backlog | 15:55 | |
| nikhil_k | if servers upgrade slower than clients, you have a faster ci/cd for client upgrades | 15:55 |
| nikhil_k | and I would assume 12 months is long time in any ci/cd (deployment strategy) | 15:55 |
| bunting | but scripts update slowly as well | 15:55 |
| nikhil_k | it can't really be called CI/CD | 15:56 |
| nikhil_k | yeah, that means the packaging being done is slower than what openstack does | 15:56 |
| flaper87 | If I read everything correctly, the current request is to make the client v2 CLI compatible with V1 | 15:56 |
| flaper87 | I scrolled through the convo really quick | 15:56 |
| jokke_ | bunting: but the script users should have preproduction before they deploy anything new knowing that their scripts work | 15:57 |
| flaper87 | is that correct? | 15:57 |
| nikhil_k | sort of but not completely | 15:57 |
| flaper87 | nikhil_k: was that with me? | 15:57 |
| flaper87 | :P | 15:57 |
| nikhil_k | flaper87: we kinda agreed to support v1 commands as hidden ones until we release 2.x.x in 2.x.x no 0.x.x commands hidden (rolled over to 1.x.x) are allowed | 15:57 |
| nikhil_k | flaper87: yes | 15:57 |
| nikhil_k | flaper87: 2.x.x will support hidden commands that have become old in 1.x.x | 15:58 |
| *** TravT is now known as TravT_ | 15:58 | |
| flaper87 | hidden as not documented and not in the `--help` output ? | 15:58 |
| *** bkopilov has quit IRC | 15:58 | |
| nikhil_k | yes | 15:58 |
| *** TravT_ is now known as TravT | 15:58 | |
| flaper87 | Would that include a deprecation warning ? | 15:58 |
| nikhil_k | and possibly some warning that let the deployers know when to upgrade | 15:58 |
| nikhil_k | flaper87: exactly | 15:58 |
| openstackgerrit | Deepti Ramakrishna proposed openstack/glance-specs: Support single disk image OVA/OVF package upload https://review.openstack.org/194868 | 15:59 |
| bunting | if they use it they would get a warning | 15:59 |
| nikhil_k | and they won't be allowed when you specific api version explicitly | 15:59 |
| flaper87 | well, that's a deprecation path :P | 15:59 |
| flaper87 | yup | 15:59 |
| nikhil_k | specify* | 15:59 |
| *** TravT_ has joined #openstack-glance | 16:00 | |
| nikhil_k | mclaren who disappeared from here :- | 16:00 |
| nikhil_k | It should be very similar to a wiki page that says something like: our default version is going to be v2, we will support v1 and have reduced support for v1 for so and so. On this date, we stop v1 support completely (by default). | 16:00 |
| nikhil_k | 11:40:36 [ nikhil_k] and in case of openstack, we can say 2.0.0 is when we stop and the tentative release period is date_range | 16:00 |
| flaper87 | to be fair, these deprecations should've been done in 0.x.x and then removed in 1.x.x | 16:01 |
| flaper87 | but anyway | 16:01 |
| nikhil_k | flaper87: I think that sounds like a reasonable leeway | 16:01 |
| nikhil_k | flaper87: in that case people would have debated the need for it | 16:01 |
| flaper87 | exactly and that'd have given us time to correct the inconvenience | 16:02 |
| flaper87 | That said, we can correct the process for 2.0 | 16:02 |
| nikhil_k | I think we are moving towards a push model when we need to push certain things to move forward | 16:02 |
| flaper87 | that is, lets deprecate in 1.x.x things that will go away in 2.0 | 16:02 |
| flaper87 | and not add in 2.0 hidden commands | 16:02 |
| * nikhil_k hopes we can get back to a pull one | 16:02 | |
| *** TravT has quit IRC | 16:03 | |
| bunting | I think its more about the default behaviour, that supports the backwards compatablitity | 16:04 |
| nikhil_k | bunting: let's see. for now we are set. | 16:04 |
| nikhil_k | we will know more on the user requests later when 2.0 is planned | 16:05 |
| nikhil_k | it's been fun. thanks guys. I will see you all around later hopefully on a different note. | 16:05 |
| flaper87 | I also think, and this is probably 2.0 material, the client shouldn't match the API 1:1. HAving a more human friendly CLI helps avoiding this incompatibilities because everything is handled in glanceclient internally rather than through CLI arguments | 16:06 |
| *** rajesht has quit IRC | 16:06 | |
| *** jistr has quit IRC | 16:06 | |
| *** jaypipes has joined #openstack-glance | 16:07 | |
| *** bkopilov has joined #openstack-glance | 16:10 | |
| jokke_ | flaper87: but we should not change our v2 API behavior just because it became default, right? | 16:11 |
| *** aix has quit IRC | 16:12 | |
| jokke_ | namely stuff like https://review.openstack.org/#/c/219802/ and https://review.openstack.org/#/c/219205/ | 16:12 |
| flaper87 | jokke_: yes | 16:13 |
| * flaper87 clicks | 16:13 | |
| *** mclaren has joined #openstack-glance | 16:15 | |
| mclaren | nikhil_k: sorry, I missed a lot of that | 16:15 |
| mclaren | can we get this https://review.openstack.org/#/c/219802 into the stable client release (in theory) | 16:16 |
| *** MarcF has quit IRC | 16:17 | |
| jokke_ | mclaren: how do you explain that backwards incompatibility to all Images API v2 users who expects Images API v2 to behave like it did day before? | 16:19 |
| nikhil_k | mclaren: maybe, https://wiki.openstack.org/wiki/StableBranch#Stable_branch_policy . it does come under the #2 user benefit criterion | 16:19 |
| nikhil_k | jokke_: so I agree. all behavior that changes the v2 defaults and makes it more like v1 is bad | 16:20 |
| mclaren | jokke_: we don't need to be backwards compatible when it comes to bugs | 16:20 |
| nikhil_k | but if we have some things that aren't in v2 and needed for v1 then it might be okay | 16:21 |
| jokke_ | mclaren: so this is not client bug but API bug discovered only now ;) | 16:21 |
| mclaren | nikhil_k: ok, thanks. So there is no principle against merging that patch, we can use our discretion. | 16:23 |
| openstackgerrit | Kairat Kushaev proposed openstack/glance: Check value format when uploading blob value https://review.openstack.org/219262 | 16:24 |
| mclaren | similarly, we can make "image-create --is-public True" work again? | 16:24 |
| flaper87 | I personally think that's a bad idea :/ | 16:24 |
| flaper87 | I even tried to do that before in glanceclient when I was working on making it work with nova | 16:25 |
| flaper87 | and the patch got rejected pretty much for the same reason | 16:25 |
| flaper87 | When are we going to remove `--is-public` ? | 16:25 |
| *** chutwig has quit IRC | 16:25 | |
| mclaren | so, we have about two weeks to potentially get things like this in. | 16:26 |
| *** sayali has quit IRC | 16:26 | |
| flaper87 | I'd like us to be very careful. Even for things that aren't in v2 because, while that will help with migration, it might end up giving the wrong message | 16:26 |
| flaper87 | Look at py2 and py3 | 16:26 |
| * flaper87 just dropped that bomb | 16:27 | |
| flaper87 | :P | 16:27 |
| jokke_ | ouuuuh | 16:27 |
| flaper87 | I'm not saying we should just go and break stuff but we won't be able to provide a 1:1 compatibility between these 2 versions | 16:27 |
| bunting | If we don't have it in the help, and show a deprecation warning, we would send a clear message? | 16:27 |
| flaper87 | and people will have to update code and scripts anyway | 16:27 |
| mclaren | If someone puts up a patch to unbreak image-create scripts with 'is-public True' why would we not do that? Why would we choose to break stuff when we don't have to? | 16:28 |
| flaper87 | Let me put it in another way: Why would we choose to maintain something that is not forward compatible ? | 16:28 |
| *** MarcF has joined #openstack-glance | 16:29 | |
| flaper87 | It's not part of the API | 16:29 |
| *** chutwig has joined #openstack-glance | 16:29 | |
| flaper87 | Note that I'm not advocating to reject everything, I'm saying lets be careful and remember that things we add we'll have to maintain and this is not solving our issue | 16:29 |
| flaper87 | this is just postponing it | 16:29 |
| mclaren | why is it only postponing it? | 16:30 |
| mclaren | (even postponing is an improvement) | 16:30 |
| flaper87 | and, FWIW, it shouldn't be news to anyone that v2 is not compatible with v1. V2 has been around since grizzly and the client has had support for it since havana ? | 16:30 |
| flaper87 | mclaren: might be, might not be. | 16:30 |
| flaper87 | we'll have to get rid of `--is-public` eventually | 16:30 |
| flaper87 | because it's not a boolean field anymore | 16:31 |
| mclaren | we can map it to visibility public, that's not hard | 16:31 |
| flaper87 | mclaren: sure but we can do the same with many other parts of our cli. | 16:32 |
| flaper87 | that goes back to what I said before. I don't think the CLI should match 1:1 our API | 16:32 |
| flaper87 | because it just makes it harder to maintain | 16:32 |
| flaper87 | For instance, do we want to talk about having to pass json through the CLI? | 16:32 |
| mclaren | ok, we're not doing it because it's hard? | 16:32 |
| flaper87 | mclaren: nope, I didn't say that | 16:32 |
| flaper87 | did I say that? | 16:32 |
| mclaren | sorry | 16:33 |
| flaper87 | I'll re-review those patches tomorrow with a fresher mind | 16:33 |
| nikhil_k | but cli should not be confusing enough for users to be look at the --debug and say they can't understand it at all | 16:33 |
| flaper87 | nikhil_k: was that with me? | 16:33 |
| nikhil_k | so, somewhere in the middle where we don't confuse users and also given them a hint that this is a deprecation period so it sorta works | 16:33 |
| flaper87 | sorry, my brain is tired | 16:34 |
| nikhil_k | yeah, same here. let's chat on this tomorrow | 16:34 |
| *** e0ne has quit IRC | 16:34 | |
| *** sayali has joined #openstack-glance | 16:35 | |
| openstackgerrit | Brian Rosmaita proposed openstack/glance: Add tasks info to glance documentation https://review.openstack.org/220166 | 16:36 |
| *** annegentle has quit IRC | 16:36 | |
| *** annegentle has joined #openstack-glance | 16:37 | |
| kragniz | mclaren, nikhil_k: another option is to make using --is-public and other options not present in v2 print a link to migration documentation | 16:38 |
| kragniz | not sure if that's been discussed | 16:38 |
| jokke_ | oh kragniz is still alive \o/ | 16:38 |
| jokke_ | kragniz: it has not and I think it's not absolutely ridiculous idea either | 16:39 |
| kragniz | jokke_: just back from the middle of the baltic! | 16:39 |
| kragniz | jokke_: it stops people from wondering why their command just broke, at least | 16:40 |
| jokke_ | kragniz: enjoyed Ahvenanmaa? | 16:40 |
| kragniz | jokke_: yeah, it was pretty nice | 16:41 |
| nikhil_k | kragniz: hey! | 16:42 |
| mclaren | kragniz: for --is-public, no -- that's easy to handle. For --copy-from, sure. | 16:42 |
| nikhil_k | kragniz: sure, that's a good idea but not sure if it helps our scripty users bunting ^^ | 16:42 |
| kragniz | jokke_: I stocked up on that salty liquorice stuff | 16:42 |
| jokke_ | uuh | 16:43 |
| jokke_ | kragniz: I really need to get there as well :( | 16:43 |
| *** smatzek has quit IRC | 16:43 | |
| kragniz | nikhil_k: good point, but it at least gives an idea on how to migrate | 16:43 |
| nikhil_k | kragniz: good help tip. we should add that to the merge proposal | 16:43 |
| bunting | nikhil_k: Yeah, got to think of the scripters ;) | 16:44 |
| jokke_ | nikhil_k: first thing I got from that was migration page ... we need to get that done as atm. is pretty much non existing | 16:44 |
| nikhil_k | jokke_: heh, I think it's the wiki that stuart setup can be renamed to indicate so. | 16:45 |
| mclaren | Cinder map back to the old style in these cases, but without showing it in the help. | 16:45 |
| jokke_ | nikhil_k: not really ... it just tels what's different, not how to get over it and continue life | 16:46 |
| nikhil_k | yeah, how to is needed for sure | 16:46 |
| nikhil_k | bunting: so that's the dependency ^ | 16:46 |
| nikhil_k | specifically 'how_to' & rename | 16:47 |
| nikhil_k | but you might get some more of these in the review | 16:47 |
| nikhil_k | thanks to jokke_ for being proactively letting know what can come up. good thing it doesn't show up on the review at the last minute! | 16:48 |
| nikhil_k | my brain is tired atm. so, I am off for lunch. catch up with you guys later | 16:49 |
| * nikhil_k & | 16:49 | |
| *** TravT has joined #openstack-glance | 16:52 | |
| *** TravT has quit IRC | 16:53 | |
| *** skylerberg has joined #openstack-glance | 16:54 | |
| bunting | I wrote this, as this is what we have seemed to come up with so far: https://etherpad.openstack.org/p/glance-backwards-compatibility | 16:54 |
| *** TravT_ has quit IRC | 16:55 | |
| *** mclaren has quit IRC | 16:57 | |
| *** harshs has joined #openstack-glance | 16:59 | |
| *** haomaiw__ has joined #openstack-glance | 17:02 | |
| *** haomaiwa_ has quit IRC | 17:03 | |
| *** spzala has quit IRC | 17:06 | |
| *** belmoreira has quit IRC | 17:07 | |
| *** spzala has joined #openstack-glance | 17:08 | |
| *** TravT has joined #openstack-glance | 17:09 | |
| *** achanda has joined #openstack-glance | 17:14 | |
| *** achanda has quit IRC | 17:21 | |
| *** julim has quit IRC | 17:21 | |
| *** achanda has joined #openstack-glance | 17:22 | |
| *** zul has quit IRC | 17:26 | |
| *** zul has joined #openstack-glance | 17:27 | |
| *** zul has quit IRC | 17:27 | |
| *** zul has joined #openstack-glance | 17:28 | |
| stevelle | sigmavirus24: I'd like to bend your ear a bit on Glance topics early next week FYI | 17:31 |
| *** annegentle has quit IRC | 17:33 | |
| *** ducttape_ has quit IRC | 17:41 | |
| *** annegentle has joined #openstack-glance | 17:42 | |
| *** e0ne has joined #openstack-glance | 17:46 | |
| *** sayali has quit IRC | 17:50 | |
| *** sayali has joined #openstack-glance | 17:53 | |
| *** julim has joined #openstack-glance | 17:58 | |
| *** ducttape_ has joined #openstack-glance | 18:03 | |
| *** sayali has quit IRC | 18:06 | |
| *** spzala has quit IRC | 18:13 | |
| sigmavirus24 | maybe | 18:25 |
| openstackgerrit | Mike Fedosin proposed openstack/glance: Disable v3 API by default https://review.openstack.org/215709 | 18:27 |
| *** harlowja has quit IRC | 18:31 | |
| *** annegentle has quit IRC | 18:31 | |
| *** harlowja has joined #openstack-glance | 18:32 | |
| *** spzala has joined #openstack-glance | 18:34 | |
| *** annegentle has joined #openstack-glance | 19:01 | |
| *** dedery has joined #openstack-glance | 19:05 | |
| openstackgerrit | Stuart McLaren proposed openstack/glance_store: Swift reseller prefix/service token support https://review.openstack.org/186769 | 19:08 |
| *** achanda has quit IRC | 19:18 | |
| *** changbl has quit IRC | 19:30 | |
| *** wokuma has joined #openstack-glance | 19:38 | |
| *** groen692 has quit IRC | 19:40 | |
| *** ayoung has quit IRC | 19:50 | |
| openstackgerrit | Mike Fedosin proposed openstack/glance: Disable v3 API by default https://review.openstack.org/215709 | 20:03 |
| *** chutwig is now known as erhud1 | 20:04 | |
| *** erhud1 is now known as erhudy1 | 20:04 | |
| *** rosmaita has quit IRC | 20:06 | |
| *** rosmaita has joined #openstack-glance | 20:08 | |
| *** akanksha has quit IRC | 20:08 | |
| *** e0ne has quit IRC | 20:12 | |
| openstackgerrit | Flavio Percoco proposed openstack/python-glanceclient: Consider `--os-token` when using v2 https://review.openstack.org/219686 | 20:16 |
| openstackgerrit | Flavio Percoco proposed openstack/python-glanceclient: Don't make `help` require auth parameters https://review.openstack.org/220096 | 20:16 |
| openstackgerrit | Flavio Percoco proposed openstack/python-glanceclient: Check if v2 is available and fallback https://review.openstack.org/219469 | 20:16 |
| *** dedery has quit IRC | 20:16 | |
| *** julim has quit IRC | 20:18 | |
| *** achanda has joined #openstack-glance | 20:18 | |
| openstackgerrit | Merged openstack/glance: Updated from global requirements https://review.openstack.org/215082 | 20:19 |
| *** julim has joined #openstack-glance | 20:21 | |
| *** achanda has quit IRC | 20:24 | |
| *** e0ne has joined #openstack-glance | 20:25 | |
| *** sdake has joined #openstack-glance | 20:25 | |
| *** e0ne has quit IRC | 20:26 | |
| *** sdake_ has quit IRC | 20:29 | |
| *** achanda has joined #openstack-glance | 20:29 | |
| *** julim has quit IRC | 20:33 | |
| *** achanda has quit IRC | 20:38 | |
| *** ayoung has joined #openstack-glance | 20:41 | |
| *** julim has joined #openstack-glance | 20:42 | |
| *** achanda has joined #openstack-glance | 20:47 | |
| *** julim has quit IRC | 20:54 | |
| *** harshs has quit IRC | 20:59 | |
| *** dramakri has joined #openstack-glance | 21:01 | |
| *** harshs has joined #openstack-glance | 21:01 | |
| *** julim has joined #openstack-glance | 21:02 | |
| *** harshs has quit IRC | 21:03 | |
| *** ayoung has quit IRC | 21:17 | |
| openstackgerrit | Tom Cocozzello proposed openstack/glance_store: Activate pep8 check that _ is imported https://review.openstack.org/220320 | 21:30 |
| *** harshs has joined #openstack-glance | 21:38 | |
| *** annegentle has quit IRC | 21:38 | |
| *** burgerk has quit IRC | 21:39 | |
| *** harshs has quit IRC | 21:46 | |
| *** ducttape_ has quit IRC | 21:46 | |
| *** ducttape_ has joined #openstack-glance | 21:50 | |
| *** haomaiwa_ has joined #openstack-glance | 21:54 | |
| *** haomaiw__ has quit IRC | 21:56 | |
| *** vijendar has quit IRC | 21:56 | |
| *** annegentle has joined #openstack-glance | 21:57 | |
| *** harshs has joined #openstack-glance | 22:01 | |
| *** annegentle has quit IRC | 22:03 | |
| *** dramakri has quit IRC | 22:09 | |
| *** kebray has quit IRC | 22:30 | |
| *** bpoulos has quit IRC | 22:43 | |
| *** harshs has quit IRC | 22:45 | |
| *** ducttape_ has quit IRC | 22:50 | |
| *** ayoung has joined #openstack-glance | 22:51 | |
| *** edmondsw has quit IRC | 22:53 | |
| *** jecarey has quit IRC | 22:59 | |
| *** sdake has quit IRC | 23:06 | |
| *** annegentle has joined #openstack-glance | 23:29 | |
| *** achanda has quit IRC | 23:31 | |
| *** dims has quit IRC | 23:44 | |
| *** changbl has joined #openstack-glance | 23:48 | |
| *** dims has joined #openstack-glance | 23:53 | |
| *** tpeoples has quit IRC | 23:56 | |
| *** aix has joined #openstack-glance | 23:58 | |
| *** TravT has quit IRC | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!