18:01:55 #startmeeting trove 18:01:55 Meeting started Wed Feb 5 18:01:55 2014 UTC and is due to finish in 60 minutes. The chair is hub_cap. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:01:56 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 18:01:59 hiall 18:01:59 The meeting name has been set to 'trove' 18:02:01 o/ 18:02:09 hi 18:02:20 oh boy 18:02:26 o/ 18:02:27 #link https://wiki.openstack.org/wiki/Meetings/TroveMeeting 18:02:33 hi all 18:02:34 o/ 18:02:34 o/ 18:02:35 hello 18:03:08 maybe we should start with updates from last meeting ? 18:03:19 o/ 18:03:22 o/ 18:03:30 denis_makogon: ? 18:03:37 tempest stuff 18:03:45 o/ 18:03:49 denis_makogon: if u want to talk about something 18:03:52 put it on the agenda 18:04:00 ok 18:04:01 #topic tempest 18:04:09 SlickNik: denis_makogon would like to know whats up w/ this 18:04:34 SlickNik, any news about image elements ? 18:05:28 denis_makogon: nope, still WIP. 18:05:35 o/ 18:05:36 oh, ok 18:05:41 I'm waiting for reviews on this https://review.openstack.org/#/c/69501/ 18:06:03 awesome SlickNik 18:06:14 im glad u didnt add /xml/ tests :) 18:06:20 denis_makogon: have you started writing other tempest tests that are not dependent on image-elements? 18:06:33 hub_cap: I'm not going to :) 18:06:51 SlickNik, i'm still learing tempest 18:06:56 I'm sure glad i spent that week figuring out how to make XML work with scheduled tasks :P 18:07:05 lol 18:07:08 lets talk about it later 18:07:09 If so, please update: https://etherpad.openstack.org/p/trove-tempest-items 18:07:33 jimbobhickville: :) 18:07:39 so moving on? 18:07:41 i'm glad i spent that time working on the file suffix fix 18:07:44 yes 18:07:47 I'm gonna ask around #openstack-qa this week to get that patch reviewed. 18:07:52 lol datsun180b 18:07:58 #topic i3 status 18:08:02 Yup, not much else to report on that. 18:08:15 #link https://launchpad.net/trove/+milestone/icehouse-3 18:08:31 so we have a few not started and a few unknown states 18:08:46 if u own one of these plz fix teh state of it, if its being worked on 18:08:55 undefined eh 18:08:56 if its not being worked on yet, talk to me, even if youve talked to me already 18:09:09 well the one i own is in review presently 18:09:17 datsun180b: privately :) 18:09:29 i wont change it right now cuz ill stop payin attn 18:10:13 ok so, if is not in i3, its not going to be reviewed or merged till after the mid cycle sprint 18:10:55 if i -2 you (or another core member fdoes) plz talk to us about it 18:10:59 if u feel it needs to be in i3 18:11:07 otehrwise we will remove -2's after i3 is cut 18:11:23 hub_cap: any way to keep reviews alive until then? 18:11:33 kevinconway: yup, click the button every 2 wks :/ 18:11:44 -_- 18:11:48 yea it sux 18:12:18 yeah, I know. :/ 18:12:20 ok any q's about it? 18:12:37 sounds good 18:12:43 got some work ahead 18:13:01 ok movin on 18:13:06 #topic mid cycle sprint 18:13:12 so weve got some great progress on this 18:13:20 yes 18:13:32 #link https://wiki.openstack.org/wiki/Trove/IcehouseCycleMeetup 18:13:37 done 18:13:39 apparently it's impossible to get a flight from ABIA to ABIA 18:13:47 if u look under event details, ther eis a schedule now 18:13:53 amy has finished the meetup 18:13:55 all you get are weird looks from the ticket agents 18:13:59 haha my bad 18:14:02 wrong window 18:14:38 lol @ datsun180b 18:14:46 You need 2 round trip tickets :P 18:14:50 datsun180b: slip a redcap $2[ 18:15:04 and we have some events that are being put on too, sponsored by parelastic, hp and rackspace 18:15:04 datsun180b: slip a redcap $20 - I'm sure he'd be happy to cart you around the airport. 18:15:27 "Where you goin?" "Here" 18:15:30 hey all - if you're a registered trove meetup participant, pls respond to hub_cap with your size by 21:00 UTC 18:15:34 t-shirt size 18:15:36 ^ ^ ^ 18:15:46 ill be coming for u if i dont have your info 18:15:53 so for our remoters, we need to come up w a solution 18:15:54 otherwise, you're getting a large by default 18:16:00 how many peeps we expectin? 18:16:02 i think maybe a google hangout w/ someone monitoring irc could ork 18:16:08 jimbobhickville: its close to 30 now 18:16:18 nice 18:16:22 in a room that fits 20~25 "comfortably" 18:16:27 so, be prepared to stand 18:16:28 hub_cap, hangout would be perfect 18:16:38 +1 18:16:57 denis_makogon: k. ill be sure to set it up 18:17:02 mattgriffin: ^ ^ 18:17:09 thanks, a lot 18:17:10 ill send the info to the ML 18:17:18 hub_cap: thanks! 18:17:28 #action hub_cap to set up hangouts for the trove meetup 18:17:31 http://static2.businessinsider.com/image/5034f4cf69beddb425000017/a-robotics-startup-is-putting-an-ipad-on-a-segway-like-roller-and-selling-it-for-2499.jpg 18:17:49 jmontemayor: Good idea 18:18:48 id also like to put together a keysigning party 18:18:53 which ive never done before 18:18:59 so it might be a bit shaky 18:19:05 yeah that sounds fun 18:19:18 hub_cap: I can help put that together. 18:19:20 o/ 18:19:53 SlickNik: ok perfect 18:20:04 Let me get a wiki page going for it. 18:20:13 SlickNik: k, i already have a good bit of info 18:20:37 I know all about gpg keysigning 18:20:39 :) 18:20:54 i know all about parties 18:21:02 lol vipul 18:21:07 it's rather trivial process assuming everyone can identify their key and you trust them 18:21:22 I think we can skip the part of the process where you show them your state/gov ID :) 18:21:22 you don't really do the keysigning at the party. 18:21:42 All you do is verify their identity, and ensure that you have the right key for them. 18:21:53 Once that's done, you can sign it at your leisure. 18:21:55 NO SIGNING AT THE PARTY! 18:21:59 NO LAPTOPS! 18:22:08 awwwww 18:22:11 boo 18:22:18 SlickNik: and i will put together some info 18:22:28 what if it is to setup a LAN party to play league of legends :) 18:22:29 and send everyone emails aobut what to do for prep 18:22:31 i've got six or seven keys i need signed at the meet up. 18:22:40 kevinconway: then u suck at pgp 18:22:41 no laptops at the meetup? 18:22:41 i'll be in the back 18:22:50 cp16net: fail... 18:22:53 lol 18:22:57 just at the key party cp16net 18:22:59 ok lets move on :) 18:23:09 kevinconway: I get this sinking feeling we'll later have to testify against you for some form of tax evasion... 18:23:26 #topic extensions management 18:23:32 #link https://docs.google.com/document/d/1vsrndOmtCwO2Cb7_6JCE6XZd1kyjWVGCBsjXeT0WOuM/edit?usp=sharing 18:23:38 denis_makogon: do u not remember that we had a HUGE discussion on this already? 18:23:41 i've sent the ML 18:23:42 and came to conclusions? 18:23:47 you participated 18:23:57 cweid was the owner and we talked abou twhat we would do 18:24:04 i remember, we came into capabilities 18:24:06 did u encorporate that discussion into your thoughts? 18:24:16 man, hub_cap, expecting people to remember things? what's up with that? 18:24:17 right and now kaleb is working on capabilities 18:24:22 jimbobhickville: lol 18:24:25 I have a comment about item #3- "each method should raise NotImplemented exception" 18:24:35 kalebpomeroy: thas kaleb 18:24:50 I think we shouldn't bother raising that. If the method isn't there, it won't be callable and will fail either way 18:24:58 yes, i researched over capabilities and found out that trove hasn't got enough flexibility to allows such things 18:25:15 I concur with grapex, the lack of the method existing should be enough of an exception 18:25:36 denis_makogon: did u talk to kalebpomeroy who is working on it? 18:25:42 because capabilities doesnt exist yet really denis_makogon 18:25:45 and its just a blueprint 18:25:51 this seems to duplicate that in a way 18:26:04 users is a capability of trove, so is volumes 18:26:22 so is root.. etc... 18:26:27 no, because i didn't knew that he's on that, and there were no ML's about trove and its capabilities 18:27:03 ya we have been discussing it at rax, hes brought it up in the channel 18:27:18 https://wiki.openstack.org/wiki/Trove/trove-capabilities 18:27:19 so lets see how those can work together, can u work w/ kalebpomeroy denis_makogon ? 18:27:27 maybe there is overlap between these 18:27:35 i suppose, channel discussions is not enough, thought 18:27:40 kalebpomeroy: is there a blueprint too? 18:27:44 denis_makogon: definitely 18:27:46 hub_cap, yes, i will 18:27:53 https://blueprints.launchpad.net/trove/+spec/capabilities 18:28:07 this was a mistake on my end, i told him to hold off cuz we were going to discuss in the trove mid cycle sprint denis_makogon 18:28:10 kalebpomeroy, lets talk about that after meeting 18:28:19 kalebpomeroy: was going to send a ML 18:28:24 denis_makogon: K 18:29:02 hub_cap, i raised this question because amcrn wrote me several times at mongo and cassandra reviews 18:29:11 so i decided to go it asap 18:29:30 denis_makogon: its fine 18:29:34 we can solve it together 18:29:45 since redis passed through, i decided to do that asap 18:30:11 ok so anythign else to add to this? 18:30:16 no 18:30:20 yall gonna chat about it after, cool 18:30:35 maybe we need ML about capabilities before meet-up 18:31:01 #topic security groups 18:31:07 #link https://docs.google.com/document/d/12zy7KYzB3e8fHjHlflU-JevN_rbkOgei4ZjGusgq1bQ/edit?usp=sharing 18:31:48 if there are no objections about oslo groups per datastore and about single ports and port ranges 18:32:11 are there any suggestions for this topic ? 18:32:19 denis_makogon: did u and I and amcrn talk about this a while ago? 18:32:21 and agree on this? 18:32:30 yes 18:32:35 im pretty sure we wantd to do nested groups, but we couldnt 18:32:42 that still doenst exist in oslo rigth? 18:32:44 *right? 18:32:51 yes 18:32:59 only stand alone groups 18:33:28 perfect 18:33:33 then i re-approve this denis_makogon :) 18:33:45 #topic backup-restore encryption 18:33:46 so, after some research over oslo and other possible ways, i found out that given implementation is more than enough 18:33:52 This looks more reasonable, with the ranges. 18:33:56 yea 18:34:00 its great work 18:34:01 #link https://docs.google.com/document/d/1cp-S_KiJjEEo_ZIgWA8lK1pLd5aWITOnnwUGEQU3yME/edit?usp=sharing 18:34:55 i'd not suggest to perform integration with BBQ, until it's not integrated into DevStack fully 18:35:12 solum followed that way and they stacked 18:35:32 denis_makogon: BBQ? 18:35:38 lol bbq 18:35:39 That sounds delicious 18:35:40 hub_cap, Barbican 18:35:41 is that barbican? 18:35:42 hahahhah 18:35:46 i made some this past weekend 18:35:48 Barbican 18:35:51 Is Barbican's nickname really BBQ? Because if so that is awesome. 18:36:15 Barbican is long enough to write its name fully 18:36:25 BBQ is enough 18:36:32 barbican is requesting incubation 18:36:35 or Barby cannon 18:36:46 barbie ;) 18:36:48 I'm getting hungry 18:36:52 lol 18:36:53 Where did the Q come from? Shouldn't it be BBC? :P 18:37:07 Barbie is a registered trademark, it will never get incubated 18:37:13 BBC, code name Dr. Who? 18:37:13 i suggest to make driver driven strategy 18:37:14 hehe lets get off the name 18:37:23 denis_makogon: sure, just like dns 18:37:27 yes 18:37:35 so, what do you think about KDF ? 18:37:36 but im not sure i see a need to do it locally 18:38:02 i researched about that almost month 18:38:19 denis_makogon: sure i mean thats good for encryption 18:38:20 denis_makogon: KDF sounds good, what's the actual value you plan to use for derivation, though? 18:38:21 fwiw 18:38:29 #link http://en.wikipedia.org/wiki/Key_derivation_function 18:38:33 but id rather u impl barbican first 18:38:40 i dont know if i want that encrytion _in_ our codebase 18:38:44 if we can use bbq 18:39:01 hub_cap: I definitely don't. I'd rather it be in barbican / oslo 18:39:03 hub_cap, since all up for driver-driven stuff than we can do that in parallel 18:39:08 no 18:39:12 not if we never need it locally 18:39:32 we will remove the rsdns driver once we use designate here 18:39:38 BBQ is only requested, RFC already released and implemented 18:39:57 denis_makogon: we dont need that extra code here 18:40:03 if the ONLY reason is that its not incubated 18:40:08 contrib 18:40:12 then thats not a good enough reason to write our own code 18:40:28 u can put it in another repo if u want but i dont see a need for it in trove if we have a whole key mgmt system being built 18:40:35 anyone agree from core? 18:40:38 SlickNik: vipul grapex ? 18:40:42 OS cryptography implemented in oslo.crypt 18:40:52 if u want to push it there, feel free denis_makogon 18:40:56 BBQ is far away from production ready PKI 18:40:59 +1 I'd rather not have to deal with a whole crypto codebase in trove. 18:41:01 hub_cap: I agree 18:41:09 and the code we write will be production ready? 18:41:09 +1 hub_cap 18:41:12 SlickNik, oslo ? 18:41:27 what about pushing it in oslo ? 18:41:29 u can guarantee that the code u write will be better than a whole proejct devoted to pki encryption? 18:41:34 and we can reuse it from there 18:41:47 that sounds like a smart approach 18:42:18 PKI allows to do more than i proposed 18:42:43 but i'm guarantee that BBQ will give us what we really need 18:42:50 I can be convinced to use PKI from oslo if Barbican's not making much traction. 18:42:58 ^^ 18:43:03 lets start with oslo 18:43:14 denis_makogon: feel free to write code in oslo 18:43:21 and make a barbican driver for us 18:43:23 i'm sure what oslo-inc will release oslo.crypto 18:43:24 But frankly we should align with what OpenStack is doing with PKI as a whole. 18:43:32 SlickNik: ++ 18:43:42 id much rather make it only bbq 18:43:51 rather than support a legacy thing for the next year or more 18:43:53 Or maybe BYOB 18:43:56 i want lunch suddenly 18:43:59 that _isint_ what the rest of openstack is now using 18:44:15 kevinconway: can u type and eat at the same time? 18:44:22 plugging system 18:44:30 don't you have 3 arms kevinconway? 18:44:35 we now have contrib dir 18:44:56 i'd rather not clutter contrib 18:45:14 denis_makogon: we can chat about it once u get your code into oslo 18:45:15 but it wont go into trove first 18:45:15 for now, if u want, u can integrate bbq into trove 18:45:15 and thats going to be in juno too 18:45:15 so its likely itll be incubated in juno too 18:45:21 denis_makogon: not in contrib 18:45:22 in oslo 18:45:25 like u said 18:45:26 ok, i'll do that 18:46:10 but still i'm not the fun of half-integrated projects 18:46:26 that is the point of openstack 18:46:30 integration w/ other projects 18:46:36 we are on the bleeding edge of these projects 18:46:38 and can help them 18:46:42 like other porjects helped us 18:46:47 its our responsibilty to do this 18:46:49 if someone read BBQ's docs - it has lots of security issues in design 18:46:51 period 18:46:57 then make it better denis_makogon 18:47:00 dont rewrite it 18:47:01 in trove 18:47:02 period 18:47:09 denis_makogon: Then we should bring it up and help fix them. 18:47:12 this is the community of together not silos 18:47:26 #action implement KDF to oslo 18:47:32 lets move on 18:47:38 good speech hub_cap 18:47:45 thx juice :) 18:48:14 what's next ? 18:48:26 #topic open discssion 18:48:33 so lets realize we have stuff in i3 18:48:42 you got to add a handle to #actions, ie #action barney steal fred's pebbles 18:48:43 and everything we hav ein i3 shoul be merged 18:48:49 lol datsun180b 18:49:33 guys patches about DB logs audit hanging almost 2 month without reviewers 18:50:12 hub_cap is lagging 18:50:13 hub_cap: is lagging 18:50:22 is hub_cap lagging? 18:50:29 la 18:50:30 what's that one about denis_makogon? 18:50:32 g 18:50:38 briefly.. 18:50:41 juice, pulling database logs 18:50:47 from image 18:50:48 from guest? 18:50:53 yes 18:50:59 for audit 18:51:12 even if we havent looked at the review yet its low on the prio (or newer than the older ones) 18:51:12 so i apoligise for not seeing all reviews yet, as does the rest of core 18:51:12 wow 18:51:12 i lagged out so bad 18:51:15 i've sent ML about that 18:51:19 wowow 18:51:36 denis_makogon: we will look at your code 18:51:37 its in i3 18:51:40 it will be merged in i3 18:51:48 ok, thanks 18:51:54 everything in i3 will be merged before i3 18:51:57 or backported into i3 18:51:58 period 18:52:07 hub_cap: you like period? 18:52:12 yup 18:52:14 today i do 18:52:21 so absolute 18:52:21 bam 18:52:28 Also, it looks like it's been failing unit tests, which makes me de-prioritize it as still WIP. 18:52:59 ++ SlickNik 18:53:07 but the problem is that sometimes we have dependent reviews 18:53:11 on projects 18:53:51 Yeah, we need to figure that one out. 18:54:12 yeah maybe some og my magic i've worked on will help with that 18:54:12 yup i end up runing them manually 18:54:27 Maybe for now leave a note in the review saying that it is dependent on another review? 18:54:28 cp16net: huuuuuuh? 18:54:38 yes everyone does that SlickNik 18:54:44 if its not in the review, then i assume it shoudl work 18:54:49 hub_cap: i've made some tricks 18:54:53 but i do have to look at all the comments from the review-ee 18:54:56 need to test it out some more 18:55:46 so we have 5 min left 18:56:06 cp16net: Let me know if you find some magic to deal with it. Let's try and automate it. 18:56:14 anything else to chat about? 18:56:28 T-shirts topic ?)) 18:56:30 strings of statuses was something mentioned yesterday 18:56:34 SlickNik: yeah we can talk about it at the meetup more 18:56:35 denis_makogon: amy mentioned it 18:56:37 already 18:56:51 then nothing else 18:56:53 pushing the status management into respective classes rather than strung throughout the code 18:57:05 denis_makogon: someone can say something still 18:57:17 open discussion doesnt need to be listed for someone to bring it up :) 18:57:29 yes 18:57:31 but if no one else has anything 18:57:43 I don't think there is much to say about it other than there are opportunities to make improvements there 18:57:44 I'm good. 18:57:50 and we can do so incrementtally 18:58:02 ok I am off to the rally 18:58:10 #endmeeting