00:02:40 <ekcs> #startmeeting congressteammeeting 00:02:41 <openstack> Meeting started Thu Jul 27 00:02:40 2017 UTC and is due to finish in 60 minutes. The chair is ekcs. Information about MeetBot at http://wiki.debian.org/MeetBot. 00:02:42 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 00:02:44 <openstack> The meeting name has been set to 'congressteammeeting' 00:03:01 <ramineni_> ekcs_: hi 00:03:08 <ekcs> topics here as usual. feel free to add/comment: https://etherpad.openstack.org/p/congress-meeting-topics 00:03:11 <ekcs> hi ramineni_ ! 00:06:08 <ekcs> ok let’s get started then. 00:06:31 <ekcs> #topc announcements 00:06:39 <ekcs> congress client 1.8 released 00:06:50 <ekcs> expected to be the final release for pike. 00:07:12 <ekcs> anything else to announce? 00:08:08 <ekcs> ok then moving on. 00:08:16 <ekcs> #topic congress server, dashboard feature freeze 00:08:38 <ekcs> pike-3 and feature freeze is due today. 00:09:07 <ekcs> I see two patches we could try and merge just in time before feature freeze release. 00:09:22 <ekcs> password encryption patch: https://review.openstack.org/#/c/486787/ 00:09:23 <patchbot> patch 486787 - congress - add encryption to secret datasource config fields 00:09:28 <ekcs> monitoring pane refresh patch: https://review.openstack.org/#/c/485455/ 00:09:28 <patchbot> patch 485455 - congress-dashboard - Primitive auto reload for monitoring pane 00:10:28 <ekcs> We could try to merge this too, but it’s a critical bug fix so not quite as urgent because it could be done after pike-3. https://review.openstack.org/#/c/486546/ 00:10:29 <patchbot> patch 486546 - congress-dashboard - Atomic rules to be displayed in UI 00:10:40 <ekcs> though it’d be nice to get it in too. 00:11:07 <ekcs> any other patches to discuss? 00:12:32 <ekcs> ok then on the first patch. password encryption patch: https://review.openstack.org/#/c/486787/ 00:12:32 <patchbot> patch 486787 - congress - add encryption to secret datasource config fields 00:12:54 <ramineni_> https://www.irccloud.com/pastebin/T9eH2zaq 00:13:23 <ramineni_> Sorry some issue with my ircclient . 00:13:32 <ekcs> do we think we can get it ready to merge? 00:14:04 <ramineni_> Yes , I'll have a quick look today and get it merge 00:14:43 <ekcs> ramineni_: great thanks! 00:14:46 <ramineni_> Tomorrow is the last day 00:14:51 <ramineni_> ? 00:15:23 <ekcs> today is last day. Thursday. right now it’s 0:15AM Thursday UTC. 00:15:46 <ramineni_> Oohk 00:16:12 <ekcs> ok let’s talk about the next patch then. 00:16:25 <ekcs> monitoring pane refresh patch: https://review.openstack.org/#/c/485455/ 00:16:26 <patchbot> patch 485455 - congress-dashboard - Primitive auto reload for monitoring pane 00:16:52 <ekcs> I don’t think I’ll be able move the checkbox. sounds simple but when I tried the style gets overridden by other horizon stuff. 00:17:00 <ekcs> will take some work and some time to figure out. 00:17:16 <ekcs> does it make sense to merge as is? 00:17:19 <ramineni_> No issue ..it's a minor thing anyway 00:17:33 <ramineni_> We could merge it and see that later 00:17:41 <ramineni_> I'll update the vote 00:18:17 <ekcs> great! 00:18:41 <ekcs> moving on to the next thing then. masahito: feel free to chime in if you have any thoughts =) 00:19:26 <ekcs> fixing horizon problem when attempting to display certain rule defined tables: https://review.openstack.org/#/c/486546/ 00:19:26 <patchbot> patch 486546 - congress-dashboard - Atomic rules to be displayed in UI 00:20:01 <ekcs> I do have one question. 00:20:11 <ramineni_> Yes 00:20:13 <ekcs> table_class_attrs[col_name] = tables.Column( 00:20:13 <ekcs> col_name, verbose_name=col_name) 00:20:32 <ekcs> 1. which name gets displayed? col_name or verbose_name? 00:20:42 <ramineni_> Verbose name 00:20:43 <ekcs> 2. is the col_name used by anything else? 00:21:26 <ramineni_> No ..for using it for internal referencing we need for that column 00:21:46 <masahito> ekcs: yes. I did a quick look to encryption patch 00:22:36 <ekcs> if verbose name is displayed, would it be safer to leave col_name the way it was rather than change it from str(i) to 'Col' + str(i) ? 00:23:00 <ekcs> I don’t know what “internal references” use it, but just in case it breaks something? 00:23:40 <ramineni_> Sure .. would do that 00:24:57 <ekcs> I also wonder if it would look better just to not display any column name. I may play around with keeping verbose_name = “” like before and see what we get. it’s nitpicking I know, but since it’s feature freeze, once we fix the critical bug we don’t get to adjust things haha. 00:25:54 <ramineni_> I Thought that .. but CLI displays as col0 col1 ..so thought it would be consistent 00:27:49 <ekcs> hmm okay. To me I expect CLI to look techie and use terms like that. where as in horizon it looks much cleaner to not throw in janky looking column names where there isn’t really intended to be any column names. 00:28:42 <ekcs> the made up column names really don’t serve any purpose except make it look weird imo. 00:29:09 <ekcs> anyway just my thought. 00:30:09 <ekcs> let me see if Ican generate an example real quick 00:30:09 <thinrichs> Hi all. Sorry I'm late 00:30:21 <ekcs> hi thinrichs 00:31:24 <ramineni_> Ok ..I can adjust that .. it's a minimal change .. if you think col names doesn't look good 00:33:24 <ekcs> ok well not a big deal either way. 00:33:45 <ekcs> masahito: sorry do you have thoughts on the encryption patch? 00:34:04 <masahito> ekcs: overall LGTM. 00:35:10 <masahito> just a minor question is the get_datasource APIs need to return decrypted config all time? 00:36:05 <ekcs> masahito: it probably doesn’t NEED to return decrypted config all the time. 00:36:24 <ekcs> hold on API let me look at that. 00:36:49 <ekcs> does it hide secret fields by default or show by default? 00:37:41 <masahito> IMO, scecret fields by default is good. 00:37:51 <masahito> I think it's ok for P-3 but it't good to change the decryption is an option in later. 00:38:33 <ekcs> I was wondering what current behavior is. because sometimes secret fields are hidden. 00:39:09 <ekcs> interesting so looking at the code it seems like datasource get_items hides secret fields. but get_item shows. 00:40:01 <ekcs> masahito: yes I guess we can consider changing API behavior to hide by default and adding a new option to show. 00:40:27 <ekcs> but yes as far as this patch, it’s addressing a separate concern (encryption in DB). 00:41:00 <ekcs> anything else regarding pike-3 patches to discuss? 00:43:04 <ekcs> ok then let’s move on. 00:44:05 <ekcs> #topic doc migration 00:44:21 <ekcs> more of just a quick note to keep everyone aware. 00:44:36 <ekcs> there’s been a big migration of openstack docs going on. 00:45:23 <ekcs> here’s the spec. https://review.openstack.org/#/c/472275/13 00:45:24 <patchbot> patch 472275 - docs-specs - Adds spec for migrating the documentation suite to... (MERGED) 00:45:45 <ekcs> I think we’ve completed the work on congress client (prioritized because final release is due). 00:46:04 <ekcs> fair amount remains on congress server that we’ll try to get done before RC1. 00:47:14 <ekcs> basically involves pulling any doc maintained outside our repo into our repo, and structuring it correctly according to the structure defined. with the goal of minimizing doc team resources needed to maintain overall doc. 00:47:23 <ekcs> an thoughts or questions? 00:49:31 <ekcs> ok moving on then. 00:49:32 <ekcs> oh and the reason it’s important is because our docs would disappear/look mangled on docs.openstack.org if we don’t complete it haha. 00:49:39 <ekcs> #topic open discussions 00:50:04 <ekcs> anything else we’d like to talk about today? 00:53:23 <ekcs> By the way I think Pike will it'll be a nice release for congress. 00:53:45 <ekcs> We'll have a small but helpful collection of library policies for users to import and play with. 00:53:46 <ekcs> We also have a new monitoring pane for people to see violations and warnings. 00:54:06 <thinrichs> That does sound like a nice release! 00:54:43 <ekcs> In addition to the many less obvious things we’ve fixed/added/refined. like asynchronous firing of actions to reduce latency. 00:54:58 <ekcs> yea great job all. 00:56:07 <thinrichs> +1 00:56:09 <ekcs> Still some work to do between now and RC1. Some I’ll have to fix some testing shortcuts taken to get things into pike-3. There’s also some catch up to do on the documentation on the new features. 00:57:46 <ekcs> 3 more minutes for any other thoughts =) 00:58:16 <qiangcao> Hi all, great features. I’ll try them. Sorry I come late. 00:58:39 <ekcs> hi qiangcao ! 00:58:54 <thinrichs> Hi qiangcao! 00:59:03 <qiangcao> hi ekcs ! thinrichs! 00:59:11 <ekcs> I should also mention it was a helpful experience for me to try and write semi-realistic policy rules. 00:59:44 <ekcs> got to have more of a user experience and could see things from the perspective of a user. 01:00:12 <ekcs> see what works really well and what some pain points are. 01:00:20 <qiangcao> any pointers to the policies rules and docs? 01:00:28 <ekcs> anyway I guess it’s time. 01:00:41 <ekcs> qiangcao: I’ll chat you on irc. 01:00:48 <qiangcao> sure! 01:00:55 <ekcs> bye all! 01:00:58 <ekcs> #endmeeting