14:00:12 #startmeeting glance 14:00:13 Meeting started Thu Apr 13 14:00:12 2017 UTC and is due to finish in 60 minutes. The chair is rosmaita. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:14 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:16 The meeting name has been set to 'glance' 14:00:24 #topic roll call 14:00:26 hello everyone 14:00:54 o/ 14:00:59 \o 14:01:26 o/ 14:02:02 let's wait one more minute 14:02:21 but hello stevelle dharinic hemanthm ... thanks for your timely attendance 14:02:31 Good Morning rosmaita 14:02:41 hola rosmaita 14:03:04 Good morning 14:03:22 hello 14:03:32 ok, guess we should get started 14:03:45 #topic updates - P-1 released 14:04:09 thanks to hemanthm , our milestone tarball is available 14:04:19 #link http://tarballs.openstack.org/glance/glance-15.0.0.0b1.tar.gz 14:04:52 #topic updates - mitaka EOL 14:05:08 #link http://lists.openstack.org/pipermail/openstack-dev/2017-April/115308.html 14:05:20 tony breeds sent out the above email 14:05:33 there are a few reviews open on stable/mitaka 14:05:40 o/ 14:05:48 jokke_: perfect timing! 14:06:04 just realized i forgot to post the agenda: 14:06:07 https://etherpad.openstack.org/p/glance-team-meeting-agenda 14:06:27 jokke_: i have a question about mitaka EOL, maybe nikhil can help on this 14:06:41 is it too late to merge stuff before EOLing it? 14:06:51 we have some open patches 14:07:00 should probably document EOL-ing stable branch also as a responsibility of release CPL? 14:07:01 just wondering if we should look them over, or just abandon them 14:07:17 hemanthm: possibly 14:07:33 though there are separate teams for release and stable branch management 14:07:36 or is it someone else's responsibility? 14:07:41 aight 14:08:15 not sure ... jokke_ , nikhil do you have any input on whether the release CPL should also handle stable branch stuff, or is it better to keep them separate? 14:08:36 rosmaita: any of those patches to stable/mitaka critical? (well, if it's critical it better not be open still) 14:08:51 hemanthm: i don't think so 14:10:07 hmm-m 14:10:21 https://review.openstack.org/#/c/416674/ might be a good one to backport, though 14:10:22 if it's not tagged yet and there is patches that are valid, sure 14:10:49 OK, then we should take a quick look at these 14:11:01 #link https://review.openstack.org/#/c/352417/ 14:11:06 that's the only glance patch open 14:12:05 storage_url one would be good to get in since it fixes the multi-tenant swift store for any basic usecase 14:12:42 let's do the glance patch first ... nikhil has a -1 on it 14:13:49 Sorry a bit distracted 14:14:04 Looking 14:14:06 sounds like the patch would have to go in, and then a follow up to refresh the config files 14:14:09 nikhil: ty 14:14:13 +1 hemanthm 14:14:32 Also on the stable branch, that's a separate responsibility 14:14:40 rosmaita: why? we can generate the sample configs in the same patch 14:14:49 Not necessarily release CPL 14:14:59 well, what i mean is someone will have to take over the patch 14:15:06 ok 14:15:08 we can't merge it as-is 14:15:14 rosmaita: I'm not convinced that missing option is worth of backporting 14:15:17 or we can, if we follow up with config refresh 14:15:22 no-one has been missing it so far 14:15:29 jokke_: i think you have convinced me 14:15:37 it's been sitting since August 14:15:43 yup 14:15:47 ok, let's look at the glance_store patches 14:15:57 #link https://review.openstack.org/#/c/416674/ 14:16:13 dharinic: you have a commnet on that that another patch should also be backported? 14:16:42 Yes rosmaita. But I guess we do not want to error out on a backport? 14:16:59 jokke_: need you to abandon 352417 when you have a minute 14:18:11 The other patch would prevent a misconfiguration by raising a glance store exception 14:18:58 probably worth doing if we're going to backport the other 14:19:46 the other glance_store patch is the config order generation 14:19:53 #link https://review.openstack.org/#/c/365059/ 14:20:15 not sure we should bother if mitaka is EOL? 14:20:27 on the other hand, i guess it wouldn't hurt 14:21:18 done 14:21:18 Yeah, we can probably skip that one 14:21:27 jokke_: do you have an opinion about the config file ordering? will just make it easier for operators to migrate config files from mitaka onward 14:21:37 the config ordering one 14:22:16 rosmaita: well one needs to take the hit anyways ... that's definitely within the scope of phase III stable branch which is security and criticals only 14:22:41 ok 14:22:45 the swift one is critical enough as it's totally breaking the service to work 14:22:51 exactly 14:23:13 jokke_: how about the swift follow-up (so you can't misconfigure)? 14:23:51 not critical ... actually having a breaking bug or just allowing one to break it 14:24:10 jokke_: please abandon https://review.openstack.org/#/c/365059/ when you have a minute 14:24:24 ok, so only https://review.openstack.org/#/c/416674/ 14:26:16 oh someone beat me to it ;) 14:26:25 jokke_: can you +2 https://review.openstack.org/#/c/416674/ after you've had a chance to look at it? 14:26:32 looking now 14:27:54 ok, while jokke_ is looking, next update 14:28:03 #topic update - next operator survey 14:28:18 ok it's not direct backport and the Mitaka one is not merged. Will dig deeper after the meeting and merge them if all good 14:28:30 jokke_: ty 14:28:39 #link https://review.openstack.org/#/c/416674/ 14:28:43 wrong link 14:28:52 #link http://lists.openstack.org/pipermail/openstack-dev/2017-April/115144.html 14:29:08 cache money 14:29:14 info about next operator survey ... i just wanted to ask if anyone wants to look it over 14:29:35 otherwise i'll just send it out on monday 14:30:16 ok, hearing no excitement, i'll just send it out on monday 14:30:27 #update forum attendance 14:30:34 oops 14:30:42 #topic update - forum attendance 14:30:53 i won't be going 14:31:05 but i believe jokke_ and nikhil will be? 14:31:09 also probaly flaper87 14:31:37 I'll be there, shame to hear that you won't rosmaita 14:32:00 yeah, i was bummed to, but rackspace decided the PTG is enough for me 14:32:12 same for me 14:32:15 ask foundation? 14:32:41 i kind of hate to do that, i think foundation money should be used for people who are independent devs working on openstack 14:32:59 plus, pretty sure i missed the deadline by about 2 months 14:33:18 better reason to ask 14:33:33 all the independents will have gotten in line in front of you 14:33:33 but it's probably worth telling the foundation that the summit/ptg split isn't working out as they had planned 14:34:07 ok, i will take stevelle 's advice and follow up with the foundation 14:34:07 Yeap 14:34:31 ok, otherwise i'll be coordinating with nikhil and jokke_ to handle the glance events 14:34:38 1 - summit (glance update) 14:34:51 2 - forum (how do you use glance?) 14:35:01 I do think that if this trend is continuing, foundation should commit to get the PTLs to the summit regardless if their company is willing to take the cost or not 14:35:09 3 - glance onboarding (during lunch one day, not sure which day though) 14:35:37 so one forum session was accepted? 14:35:51 jokke_: yes, they asked us to combine the 2 into 1 14:36:07 ah 14:36:12 i'll take a stab at a combo and send it to you and nikhil for comments 14:36:23 sure 14:36:32 1 and 2 are back-to-back, on monday 14:36:42 apparently only 3 rooms away from each other 14:36:50 that is decent 14:36:56 ok, that's all the updates 14:37:01 #topic virtual midcycle 14:37:25 the schedule has it next week ... wondering if we should hold off for one more week? 14:37:41 why delay? 14:37:51 just because we don't have the date/time yet 14:38:08 Yeah 14:38:09 or an agenda 14:38:30 are either of those going to change if we put it off? :) 14:38:37 possibly 14:38:55 though on the other hand, possibly not 14:39:17 ok, is there anyone here who CANNOT attend next week? 14:39:43 It may help me to delay unless it's Friday 14:39:48 (and the next question is: is there any day next week that is absolutely no good) 14:40:07 how is friday for people? 14:40:09 Mon Tue are no go for me 14:40:24 and sounds like wed thurs no go for nikhil 14:40:31 Fri is ok if you guys are willing to start early so it won't run until night for me 14:40:53 like 12:00 UTC? 14:40:59 I have reoccuring social at Fri evenings 14:41:03 that is going to be rough for stevelle 14:41:20 well as long as we're done by ~1700 UTC I'm fine 14:41:38 but that's just Fri 14:41:46 Checking Thursday 14:42:15 ok, let's tentatively say: Friday April 21 between 12:00-17:00 UTC (not necessarily that long, just during those hours) 14:42:30 Ok I can do Thursday 14:42:42 looking my calendar, Thu would probably be the best day any week for this 14:43:07 ok, anyone else CANNOT do Thursday April 20 ? 14:43:08 But maybe have to sneak out in the middle 14:43:39 jokke_: what time range works for thursday for you? 14:43:57 rosmaita: Thu is fine ... I can adjust 14:44:26 ok, i will send out an announcement later today with an etherpad link to fill in time slots with topics 14:44:29 that implies cancelling meeting that week? 14:44:34 stevelle: yes 14:44:51 ok, 5am it is :P 14:45:09 ok, i will skip item 3 14:45:09 Lol 14:45:13 stevelle: Thursday we don't need to do it that early 14:45:37 stevelle: let's say half five :P 14:45:38 I doubt if that time was a suggestion 14:45:48 one thing i forgot: i need volunteers to help test out the video 14:46:01 need someone from europe and someone from APAC 14:46:07 seems to work fine in USA 14:46:28 but, if anyone else wants to try it out, that's fine, too 14:46:52 please shoot me an email and i'll send you a link and we can connect and test 14:47:02 ok, let's get to the specs 14:47:06 #topic specs 14:47:18 freeze is tomorrow at 23:59 UTC 14:47:33 #link https://etherpad.openstack.org/p/glance-pike-specs-review 14:47:55 ok, taking it from the top 14:48:16 the sticking point for the 'protected' image-list filter is whether it should be strict or permissive 14:48:40 i did quick research, nova has a permissive filter for 2 booleans for the server list call 14:49:08 so, i'm inclined to say let's go permissive, unless there's a strong contrary opinion 14:50:00 there's -1 strength contrary opinion currently 14:51:37 I just really don't like service quessworking the API input 14:52:10 I'd like to bring that French example on the table, should we then allow all the translations of those workds as well? 14:52:23 -k 14:52:32 you could put a patch to oslo_strutil 14:52:37 might be a good idea! 14:52:48 anyway, if we use oslo_strutil, we'll be future proofed! 14:53:10 rosmaita: I kind of like to disagree again :P 14:53:32 i guess the problem is that since nova is permissive, consumers will assume we are, too 14:53:39 Just pushing the responsibility to someone else by using their lib doesn't mean we would not be breaking our api if something changes there 14:54:14 well, it wouldn't be a breaking change 14:54:26 all the old stuff would work correctly, plus some new stuff 14:54:42 the docs will just say that 'protected' is a boolean 14:54:58 so i imagine that most people will use 'true' or 'false' in some case combination 14:55:15 anyway, let's continue this on the patch, or in #glance 14:55:22 sure 14:55:39 nikhil: wonder what you think of the suggestion to postpone https://review.openstack.org/#/c/450621/ ? 14:55:50 Yeah 14:56:12 I don't think the discussion is possible too in pike 14:56:23 i think let's postpone removal to Queens, and possibly follow jokke_ 's suggestion and never remove it 14:56:40 jokke_: flaper87 and I can chat a bit at the summit 14:56:46 excellent 14:56:58 I have been with what jokke_ is saying 14:57:13 Just need to get that out of flaper87 system 14:57:18 :) 14:57:36 ok, https://review.openstack.org/#/c/206120/ could use some love 14:57:46 nikhil: we just bring some Tequila in and flaper87 will be sorted ;) 14:57:55 Ha 14:57:59 i think it's a good idea, but i'm on the line about the config option 14:58:00 Done 14:58:35 shoot, we're about out of time 14:58:42 sorry about the bad time management today 14:59:08 please make it a priority to take a look at the remaining specs today 14:59:30 thanks everyone! 14:59:38 thanks 14:59:58 #endmeeting