17:00:22 #startmeeting Designate 17:00:27 Thanks everyone! 17:00:27 Meeting started Wed Mar 12 17:00:22 2014 UTC and is due to finish in 60 minutes. The chair is kiall. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:00:28 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 17:00:31 The meeting name has been set to 'designate' 17:00:35 Hey Guys - Who's around today? 17:00:37 here 17:00:41 o/ 17:00:47 #link https://wiki.openstack.org/wiki/Meetings/Designate agenda 17:01:03 here 17:01:27 Cool - I believe mugsie is on his way 17:01:31 #topic Review action items from last week 17:01:50 o/ 17:02:06 present 17:02:08 all to file any remaining known bugs and target anything important to i3 - I believe we managed this, getting a couple of last minute bugs ironed out.. So - Calling it done 17:02:08 here 17:02:28 Next was: betsy to write up BP for table per record-type change 17:02:43 Did you find the time betsy? :) 17:02:49 betsy running late 17:02:50 She's not here right now…I think she did that though... 17:03:18 tsimmons: ok - cool, I didn't notice the BP. Do we have a link handy? 17:03:28 looking now... 17:04:31 Well maybe not... 17:04:33 A quick look at the wiki+blueprints lists don't show it. No worries, we'll come back next week (or at the end if someone finds it!) 17:04:39 Alright. 17:04:57 Next up, from the previous meet, was MiniDNS - kiall to refresh the spec, split blueprints, and get that POC code together. 17:05:11 I'm going to have to apologize again - this still isn't done. 17:05:30 Suffice to say HP has been keeping myself / mugsie / ekarlso busier than usual. 17:06:31 So - I'll have to look at this again over the weekend, there's literally no chance I'll be able to this week :( 17:07:11 And - That was the last un-closed action. Moving on before anyone has time to beat me ;) 17:07:21 #topic Icehouse 3 Release 17:07:30 kiall: are we still going to be able to sync up tho? 17:07:36 before the weekend on the POC 17:07:38 ? 17:07:58 the Central-API stub ? 17:08:20 eankutse: yes - we can grab an hour somewhere - let me get back to you at the end and we can find a spot that works for both of us! 17:08:30 cool:-) 17:08:45 Re i3: A final couple of fixes landed earlier today, has anyone got any final blockers before I `git push` the i3 tag? :) 17:09:21 After this, we'll have 1+ RC's before the final Icehouse release. 17:09:37 i need to investigate one 17:09:51 there was one issue with postgres and PowerDNS 17:09:53 it appears that we can set the id while doing a post 17:10:01 not sure if we should count it as a blocker 17:10:05 is this desirable? 17:10:18 vinod: Oh - Humm. That's not desirable at all. Only with V2? 17:10:35 i tested it out with v2 17:10:40 i did not yet try it out with v1 17:11:10 vinod: Okay, I'll hold off the tag until tomorrow - Hopefully we can get a fix before then? 17:11:14 #link https://bugs.launchpad.net/designate/+bug/1289444 <- effectivly blocks anyone from using PowerDNS and postgres 17:11:32 should we prioritise that as well? 17:11:36 i will work on it right after this meeting 17:11:37 I tried just now (on v1), got a 400 bad request, but I just changed one of the UUID characters, so it may have been invalid. 17:11:52 mugsie: I think we can fix that after i3 - We're not testing the PowerDNS migrations etc against Postgres automatically 17:12:17 ok, cool 17:12:18 tsimmons: so long as it was changed to something in [a-f0-9], it would still be valid 17:12:35 vinod: is there a Bug # for the issue you found? 17:12:42 yeah I just changed an "a" to a "b" 17:12:58 not yet - i wanted to make sure it was an issue - before filing it 17:13:12 Makes sense :) 17:13:16 i.e. not just an issue in my environment 17:13:22 It's def an issue! Setting the ID is asking for issues 17:13:37 Allowing the user to set the ID that is 17:13:50 there is currently 4 other bugs targeted for i3 17:13:53 #link https://launchpad.net/designate/+milestone/icehouse-3 17:14:19 we going to say these are all pushed to RC1? 17:14:25 3 now ;) 17:14:52 prev links aren't included at all right now (same as with nova's etc APIs). Pushed to Juno as it's too late for more features! 17:15:15 How about bugs #1281837 and #1288190? 17:15:33 https://bugs.launchpad.net/designate/+bug/1281837 should technically be pushed too - But it would be an API change (a change it what gets returned). I'd LOVE to get that out before we release v2 into the wild 17:15:47 a change in what gets returned* 17:15:48 we can push it to RC1 though 17:16:15 mugsie: yea - exactly, even if it's technically a feature addition ;) 17:16:39 1288190 - has this been confirmed by anyone? 17:16:48 and 1288190 is def a bug, but it's a really rarely used feature. So - I don't think it blocks rc3 17:16:53 mugsie: I've eyeballed the code, and it's a bug 17:17:14 rc3? or i3? 17:17:24 sorry - i3 17:17:29 cool 17:19:04 #link https://blueprints.launchpad.net/designate/+spec/api-version-2 17:19:09 Okay - Any others we think need to be addressed in i3? 17:19:13 is ^ in a place to mark as complete? 17:20:00 Yea - We have all the required pieces in now, some planned new additions are the only things left which don't break the API conpat 17:20:02 compat* 17:20:05 bugs wise, only the one vinod found 17:20:44 Okay - So - Just 1 blocking i3 then 17:20:59 yup 17:21:10 Moving on then 17:21:27 #topic Server Pools (Status, Getting Started) (from tsimmons) 17:21:37 Basically I wanted to get a feeling for where Server Pools are right now, what the next steps are, and how we can help. 17:22:24 I need to complete the bp (even make them basic high level info) 17:22:33 I am planing on having some time this evening 17:22:51 then whoever is interested, we can split up the work 17:23:01 I expect we can get started on actually writing code for it + minidns pretty soon, once i3 is out and the BP's have been refreshed there should be components we can start splitting out 17:23:31 i know Rackspace have some bandwidth coming up...whats the time scales for you guys? 17:24:56 Depending on how much work is available, we should be able to get some resources on it over the next few weeks. Personally I'm ready to go right now, I think some of the other devs will have some time in the next couple of weeks. 17:26:15 tsimmons: cool! I know at HP we're pretty tight on available time right now between a couple of internal things 17:26:32 Once those clear up, we should have MUCH more time to actually write code again 17:26:45 We are particularly interested in pools+miniDNS as it gets us to the robustified Bind deploy. 17:27:08 I am available for miniDNS now 17:27:15 jmcbride: Yea - I guessed as much :) I also think they both go hand in hand in terms of dev work 17:27:35 It's definitely important work that we'd love to be a part of :) 17:27:44 I was going to leave this comment for next week's week - but - since we're kinda on it now anyway... 17:27:55 so getting mugsie's vision in a BP will be the seed we look forward to in knocking it out. 17:27:59 yup 17:28:14 I'd like to Juno all about getting BIND9 working 1000x better, Pools, MiniDNS, and virtually nothing else 17:28:44 Those are all big chunks of interrelated work that will go a long way towards fixing up lots of our issues! 17:29:03 yep 17:29:05 We're all for that. 17:29:07 agreed 17:29:14 I'd like to get the IPA backend working by Juno as well 17:29:31 but I don't think that will affect any of bind9/pools/minidns 17:29:50 richm: Depending on how that works, it'll most likely benefit from the same stuff :) 17:29:57 yes, most likely 17:31:14 Okay - Well, mugsie, why don't you try and get the pools BP's updated before Monday? (which is about the soonest I can have my MiniDNS ones ready) 17:31:23 yup 17:31:29 action me on that ;) 17:31:35 That leaves 2 days for people to read + comment, and we discussed them at next weeks meet? 17:31:50 #action mugsie to refresh Pools BP's before Monday 17:31:56 good goals 17:32:01 #action kiall to refresh MiniDNS BP's/Specs before Monday 17:32:10 mugsie: please don't sacrifice your evenings and any beer drinking time ;) 17:32:29 jmcbride: theres not much chance of that ;) 17:32:35 jmcbride: So, don't do it during working hours then? ;) 17:32:44 ;) 17:33:11 #action add agenda item for Pool/MDNS BP/spec discussions 17:33:54 Okay .. 17:33:58 #topic Open Discussion 17:34:15 eankutse: checking my calendar now.. 2 mins 17:34:20 k 17:34:21 still want to talk to you about fixed IP PTRs at some point. 17:34:24 anyone have anything else while I'm doing that? 17:34:48 kiall: you can email me 17:34:54 I have my updated on the Records table 17:35:00 Gah - 9 new emails since the beginning of this meet. Never ending :'( 17:35:17 I did make a placeholder blueprint to separate the records table into multiple tables 17:35:29 Still trying to meet up with our DBA to ask him questions about it 17:35:37 That's all I've got right now 17:36:50 eankutse: Sure - I'll email you in the next few mins then 17:37:04 betsy: cool - I didn't see anything that sounded like it, what did you call it? 17:37:04 k 17:37:32 I think the short name was record-table-change 17:37:45 Should have been something like records-table-redesign 17:38:09 aha - https://blueprints.launchpad.net/designate/+spec/records-table-change 17:38:41 yep. Mostly just a place holder at this time until I meet up with the DBA to discuss in more detail 17:38:56 Cool - Hopefully we can figure out if there's a terrible flaw in the plan before next week.. But I expect there won't be ;) 17:39:03 Then I'll write up a spec and we can discuss 17:39:14 Cool 17:39:33 Just FYI, it looks like you can set an ID in v1. 17:39:58 and a created_at and updated_at too. 17:40:05 tsimmons: Humm - OK. That's a bug, one that was hopefully introduced not too long ago! 17:40:07 and created_at and updated_at fields too depending on the database 17:40:30 filing a bug now 17:40:41 Yeah. I just verified that bug 17:40:45 Could you guys include that in the bug? We'll have to figure out exactly why that's happening. 17:40:48 … in v1 17:41:05 will do 17:41:11 Thanks 17:41:27 So it's a bug in v1 and v2 17:41:34 Do they need separate bugs or just one bug? 17:41:43 1 should do 17:41:43 I'm assuming one 17:42:00 betsy: without looking, I would hazard a guess the bug was introduced as part of the V2 additions 17:42:14 But.. We'll find out soon enough 17:42:54 I don't think I ever thought to test that before 17:43:06 But makes sense it was introduced in v2 17:43:25 betsy: I'm pretty confident HP has tested for that in the past during security reviews etc. 17:43:43 Kiall: Oh, that's right. Y'all are actually running it in production 17:44:04 betsy: right - Hence I'll be digging into this the moment we're done ;) 17:44:28 Okay - So anyone have any other topics to discuss? 17:44:32 nope 17:44:37 question about backends - is create_tsigkey/update_tsigkey supposed to do anything other than create/update a TSIG RR? 17:44:38 Not from me. 17:45:50 richm: the TSIG key support today is intended to create the tsig key config (not record) in the backend, and allow that key to AXFR all zones 17:46:25 It's primarily there to allow Akamai to AXFR from our PowerDNS instances without leaving it open for anyone to AXFR (or constantly updating IP allow lists) 17:47:01 It should be extended to cover off other use cases for TSIG keys, e.g. a user should be able to create a key that allows AXFR of just their zones etc 17:47:56 i have code written for https://wiki.openstack.org/wiki/Designate/Blueprints/ReverseFixedIP. i'm testing it out. still would like to discuss the implementation. 17:48:41 Cool - is it up for review? 17:49:09 we looked at it last week. i'd still like to review the BP and make any changes we want. 17:49:25 we were suppose to meet on Friday, but you were busy. 17:49:56 would it be better for me to submit the code before the BP is approved? 17:50:04 Apologies :( Busy would be an understatement.. 17:50:15 rjrjr__: You could submit it as a WIP 17:50:35 Sure - the code can often clear up any questions that might come from the BP anyway :) 17:51:00 okay. is there something i need to put in the comment section to link it to the BP? 17:52:05 Was the BP updated recently BTW? Looks different to what I remember looking at last night! 17:53:07 yes. it is more reflective of the implementation. i went with a call to the Neutron list ports. 17:53:13 rjrjr__: in the commit message, include "blueprint fixed-ip-ptrs" and it should link things together 17:53:21 (on it's own line at the end) 17:53:54 rjrjr__: And prefix the commit with "WIP" if you'd like to submit it as a work in progress. 17:54:13 originally, i was going to use the Nova API, but i think there is an ORM bug in it that prevents me from getting the VM instance for an IP address. since the Neutron "plumbing" was in place, i decided to use the list ports call. 17:54:46 rjrjr__: I think that's probably the best way anyway, as Nova's network APIs will (eventually) be removed.. 17:55:15 my concern is not having an ID to key off of. i'm using the region and address itself. 17:55:32 the code will make this clearer. :) 17:55:54 rjrjr__: yea, agreed. I'm just unsure how we handle overlapping networks in that case :( 17:56:17 Anyway rjrjr__ I know I'll have time to eyeball the change tonight if it's up .. Hopefully others will too 17:56:33 3 mins before the Trove folks arrive with pitchforks, any final topics? 17:56:41 well, my testing is promising, so in a couple of hours i'll submit the code for review. 17:56:50 Opened bug 1291518 to track the id issue. 17:56:54 #link https://bugs.launchpad.net/designate/+bug/1291518 17:57:04 rjrjr__:awesome 17:57:20 cool :) 17:57:40 Okay - Thanks all 17:58:04 vinod: I'll ping you in a few re that bug - looking into it now. 17:58:05 #endmeeting