16:19:03 #startmeeting oslo 16:19:04 Meeting started Mon Sep 14 16:19:03 2015 UTC and is due to finish in 60 minutes. The chair is dims. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:19:06 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:19:08 The meeting name has been set to 'oslo' 16:19:16 anyone still here? :) 16:19:18 courtesy ping for GheRivero, amotoki, amrith, bknudson, bnemec, dansmith, dhellmann, dougwig, e0ne, flaper87, garyk, harlowja, haypo, 16:19:18 o/ 16:19:18 courtesy ping for ihrachyshka, jd__, jecarey, johnsom, jungleboyj, kgiusti, kragniz, lifeless, lintan, ozamiatin, redrobot, rpodolyaka, 16:19:18 courtesy ping for sergmelikyan, sreshetnyak, sileht, sreshetnyak, stevemar, therve, thinrichs, toabctl, viktors, zhiyan, zzzeek 16:19:20 o/ 16:19:20 supppp 16:19:22 o/ 16:19:25 ./ 16:19:28 o/ 16:19:29 o/ 16:19:37 hi 16:19:42 #topic Releases for Liberty 16:19:44 o/ 16:19:58 o/ 16:20:01 stable/liberty branches are cut for all oslo projects. yay 16:20:10 \o/ 16:20:24 so go on and +2/+A things in review please 16:20:39 yaaa 16:20:42 cool 16:20:54 if we need something from master to stable/liberty, please ping me 16:21:00 #topic Red flags for/from liaisons 16:21:15 any liaisons still around? :) 16:21:22 yes and nothing to report. 16:21:33 the gate ... 16:21:34 nothing for keystone. 16:21:35 hi 16:21:42 we've been only doing g-r and translations for some time now...so hoping all's well 16:21:57 thanks amrith bknudson 16:22:23 Sorry, I am kind-of here. 16:22:29 hi jungleboyj1 16:22:30 No red flags from Cinder. 16:22:43 thanks! 16:22:43 We are just in bug fix mode now. 16:23:11 Thank you dims... Sorry for being distracted tdoday. 16:23:41 fix all the bugs! 16:23:42 no worries jungleboyj1, i reached here late as well 16:23:50 #topic virtual doc sprint 16:24:01 totally got the timing wrong on the sprint 16:24:16 does anyone have time next thu and friday? 16:24:24 me 16:24:43 next thu, not this one right? 16:24:44 24 and 25th sept 16:24:46 k 16:24:47 y 16:25:09 I've got nothing going on next week. 16:25:18 bknudson: awesome 16:25:19 :) 16:25:33 and the gate is definitely not going to be broken 16:25:37 right 16:25:50 I can help 16:25:55 the gate is never broken, idk what u guys are talking about 16:25:58 lol 16:26:09 jecarey: awesome, do you want to coordinate? 16:26:53 someone needs to supply chips and dip. 16:27:09 sure ... do we have specific libraries we want to focus on? 16:27:10 * dims volunteers for chips and dip 16:27:17 virtual chips and dip? 16:27:54 jecarey: don't know yet, we can use the etherpad to figure out 16:27:57 #link https://etherpad.openstack.org/p/oslo-liberty-virtual-doc-sprint 16:28:26 jecarey: need to send an email out with the date too 16:28:59 #topic bugs, bugs, bugs 16:29:14 have to ask, do we have things we really need for liberty? 16:29:33 i saw some oslo.db stuff for logging tracebacks 16:29:37 in review 16:29:45 https://review.openstack.org/#/c/221606/ would be sorta nice, idk 16:29:50 there was a review for hiding tokens in logging 16:29:51 but idk if its 'really need' 16:29:55 right 16:30:07 not doing sorta-nice :) 16:30:12 :-P 16:30:48 dims: I missed the oslo.db one, will take a look 16:31:02 #action jecarey to send out email for virtual sprint :) 16:31:11 ok jecarey? :) 16:31:28 thanks rpodolyaka1 16:31:33 yes 16:31:56 #topic Summit planning 16:32:17 harlowja_at_home: can you take point on this please? https://etherpad.openstack.org/p/mitaka-oslo-summit-planning 16:32:34 what new libraries we need? drivers? 16:32:36 sureeee 16:32:50 i saw a blueprint for kafka with oslo.messaging for example 16:32:54 right 16:33:04 kafka def requires a session i think 16:33:08 we'll need a blueprint to nuke oslo-incubator 16:33:35 rpodolyaka1: any new stuff in oslo.db? 16:33:57 amrith: kzaitsev_mb: any new fun stuff we can do in oslo? 16:34:22 dims, none that I have right now from the trove team 16:34:37 dims: I've filed one https://blueprints.launchpad.net/oslo.messaging/+spec/messaging-api-ext but spec is in progress 16:34:39 we're looking into encryption of amqp traffic with o.m 16:34:40 oh, over the weekend found this fun looking middlware thingy - https://github.com/racker/eom/pulse 16:34:46 haven't enabled that by default yet 16:34:55 but are trying to make sure those things all work as expected. 16:35:05 you may get some requests out of that. 16:35:09 amrith: transport level or payload level amrith? 16:35:12 dims: probably just additional work based on the EngineFacade merged in Liberty (like automatic retries on deadlocks, etc) 16:35:19 ozamiatin: cool 16:35:25 dims: not sure it's big enough to have a bp/spec 16:35:55 dims, transport level. 16:35:56 * rpodolyaka1 needs to a write a blog post to 'advertise' EngineFacade 16:36:10 kgiusti: johnsom: bknudson: anything you can think ok? 16:36:14 rpodolyaka1: ++ 16:36:39 rpodolyaka1: reminds me, haven't seen zzzeeek in a while 16:36:39 I don't have anything 16:36:41 amrith, i added something to https://etherpad.openstack.org/p/mitaka-oslo-summit-planning (it'd be really nice to finally get something going/finished for trusted messaging stuff in o.m) 16:36:42 are you running bandit? 16:37:00 amrith, maybe at summit we can really figure that out (finally) 16:37:10 could talk about obfuscating / encrypting config options 16:37:30 harlowja_at_home, sure 16:37:31 also could talk about supporting replacements 16:37:36 dims: yeah, me neither, he must have been busy working on something else 16:38:00 harlowja_at_home: the privsep spec will need a session 16:38:03 rpodolyaka1: ack 16:38:11 dims, def 16:38:23 bknudson: ++ to safe password storage 16:39:03 Yeah, I like that one 16:39:12 bknudson: need a session to advertize the new oslo.policy stuff? 16:39:16 bknudson, ok i added '* safe config password storage (aka, the whattt my passwords are in clear-text, woah), keeping somethings in a vault... (or better name)' 16:39:23 feel free to rename, seems like a useful session 16:39:51 dims: I'm sure there will be a keystone session on policy (aka dynamic policies) 16:39:58 big ticket...is there anything we can do to help python3 adoption? 16:40:03 bknudson: ++ 16:40:26 * dims invokes haypo 16:41:31 ozamiatin: you need one session for zmq right? 16:41:45 dims: right 16:41:58 ozamiatin: please update https://etherpad.openstack.org/p/mitaka-oslo-summit-planning 16:42:13 dims: ok, will do 16:42:18 thanks 16:42:58 sorry, tuned out for a second reading what bandit was. Why is OSLO storing passwords? is this something new? 16:43:14 amrith: oslo.config files have passwords right now 16:43:21 yes 16:43:34 i don't think oslo is, oslo.config is (sadly imho) and it'd be nicer to have some kind of let's call it oslo.config.vault or something 16:43:38 we need to figure out how to take them out of there 16:43:39 and the passwords are easy to read now ;) 16:43:49 exactly :) 16:43:50 isn't that something to push into barbican? 16:43:55 something like a oslo.config.vault would fix that (maybe?) 16:44:02 harlowja_at_home: if we put them in a vault, where would we store the pass phrase to open the vault? 16:44:12 * amrith runs off to look at oslo.config.vault 16:44:22 amrith: there isn't one :) 16:44:23 amrith: if we put them in barbican, where do we put the credentials to get into barbican? 16:44:24 :-P 16:44:30 haha 16:45:06 what do other frameworks, services do for their passwords is a better leading question i think 16:45:08 so knowing what yahoo has done for this, i believe there is an easily rotatable key that is used to get into the 'vault' 16:45:24 but i can try to find people to comment on the yahoo internal solution for just this problem.. 16:45:51 harlowja_at_home: that would be good 16:45:55 def 16:45:58 dhellmann: i was asking before, is there anything in oslo we can do around python3 to help other projects 16:46:09 dims: sorry, distracted by release work today 16:46:16 no worries dhellmann 16:46:32 I'm not sure there's anything in oslo to do, but if someone wanted to take over my devstack patch next cycle that might be useful 16:46:40 https://review.openstack.org/181165 16:46:51 I'm going to try again, but time... 16:47:04 dhellmann: ack. i can take that over 16:47:11 dhellmann, this is a catch-22 16:47:58 amrith: right. We've always viewed oslo.config as the lowest-level for managing this sort of info. 16:48:33 it's the same reason we don't create a service for distributing oslo.config content instead of using files -- we would need a config file to tell an app where to talk to the config service 16:48:55 there's already services for distributing config files 16:49:03 right 16:49:40 from time to time we get folks suggesting that oslo.config accept a URL instead of a filename, so it can hit a remote service. But how do you tell an app what the URL is? 16:50:40 * dims peeks at openstack-specs to see if there's things we can do from there - https://review.openstack.org/#/q/status:open+project:openstack/openstack-specs,n,z 16:50:42 metaconfig :-P 16:50:46 oslo.metaconfig, ha 16:50:59 I'm not saying it's a bad idea, just that none of the solutions proposed so far have made things "simpler" in any appreciable way. 16:51:30 "asynchronous user event support to OpenStack" 16:52:36 quota's - everyone needs them and everyone has something right now 16:53:14 dims: the conclusion we reached on a centralized quota thing was it needed to be a service, since it needs a database schema 16:53:17 dims, i agree to that, i have some coworkers working on that, i'm not sure about oslo.quotas yet 16:53:30 oslo.quotas or something maybe someday 16:53:35 that service is not keystone 16:53:38 I can't remember if that was paris or vancouver where that discussion happened 16:53:40 could be useful to brainstorm though, idk 16:53:41 looks like amrith had to drop off, i remember he was talking about hiding things in nova that are started by trove from end user 16:53:50 no, there's an existing thing that starts with B that needs to be fixed up and hardened, I think 16:54:09 boson, meh, idk 16:54:27 harlowja_at_home: yeah, that's the one 16:54:44 I haven't looked at the code or anything, so I don't know if it's what we'd want or if we'd want to build something else 16:55:02 to me quotas really require a much larger initiative across all projects to get fully resolved, folks i know are working on pieces of it, but ya, time, people, cross-projects need to all really agree... 16:55:24 right 16:55:29 so i'm really thinking, ummm, maybe in a couple of years, idk 16:55:47 k that's all the ideas i had in my head :) 16:55:54 unless some TC group wants to help out really push this :-P 16:55:58 dims: it would be interesting to see if there's some work in the WSGI app level that could be pulled into oslo. Some of the flask/pecan setup stuff, for example. 16:56:16 dhellmann: interesting 16:56:51 dims: an oslo.rest lib maybe? with different parts for a flask app and a pecan app 16:56:57 dhellmann: there was something that the docs group (annegentle) was trying/thinking in terms of documenting services as well 16:57:02 related to that 16:57:10 let's spend the time getting stuff done 16:57:22 yeah, I've had pretty good luck with adding sphinxext modules to a couple of libraries for doc work 16:57:44 k. we can talk more next meeting 16:57:50 i have one more topic 16:57:57 bknudson: ++ -- it would be good to work on the lib docs, if we don't sprint before then 16:58:06 #topic next ptl 16:58:08 we're sprinting next week 16:58:11 * dhellmann realizes ++ with -- is confusing 16:58:13 oh! 16:58:22 bknudson, for quotas, i think its getting stuff done + a crap-ton of getting projects to agree, both need to happen (and probably at the same time) 16:58:24 * dhellmann checks his calendar 16:58:27 dhellmann: y picked next thu/fri 16:58:44 ++ 16:59:58 so, i have fuel in the tank for one more round (M) as ptl. but would be great to see a vote so would encourage anyone else who wishes to run as well 17:00:15 so are u half full or half empty? 17:00:16 lol 17:00:25 i believe we have this week to file nominations 17:00:28 harlowja_at_home: haha 17:00:33 it would also be good for anyone who thinks they want to do it to shadow dims this cycle (assuming he is re-elected) to learn the ropes 17:00:34 harlowja_at_home: half full 17:00:39 1000++ 17:00:59 so please think about it 17:01:10 almost out of time 17:01:33 thanks everyone. talk to you next week. apologies for the delayed start 17:01:49 #endmeeting