17:00:16 <kiall> #startmeeting designate 17:00:17 <openstack> Meeting started Wed Jan 22 17:00:16 2014 UTC and is due to finish in 60 minutes. The chair is kiall. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:00:18 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 17:00:20 <openstack> The meeting name has been set to 'designate' 17:00:31 <artom> o/ 17:00:32 <kiall> Hey all - Who's about today? 17:00:36 <vinod1> here 17:00:38 <betsy> o/ 17:00:46 <mugsie> o/ 17:00:52 <kiall> Cool :) 17:00:54 <eankutse> here 17:01:29 <jmcbride> Good day folks! 17:01:54 <kiall> So - We have no pre-prepared agenda for today, With the mini-summit so close, I've been full steam ahead on some internal work. Trying to complete a good chunk of it before travelling 17:02:09 <mugsie> same ^ 17:02:17 <kiall> #topic Review action items from last week 17:02:38 <kiall> so, we had 1 item from last .. "send email to Openstack mailing list" 17:02:44 <kiall> that was re the mini-summit 17:03:08 <kiall> jmcbride sent that during the week 17:03:08 <kiall> #link http://lists.openstack.org/pipermail/openstack-dev/2014-January/024862.html 17:03:25 <mugsie> any extra interest form that jmcbride ? 17:03:29 <kiall> jmcbride: did you get any response privately? or via eventbrite? 17:04:21 <jmcbride> so, mostly Redhat folks are now trickling in 17:04:26 <artom> I said I'd attend via Hangouts... 17:04:27 <jmcbride> but otherwise, I have not seen anything 17:04:38 <artom> Not sure how that will work out... 17:04:58 <jmcbride> I'll check our current registration 17:05:18 <kiall> artom: we'll be as inclusive as possible :) 17:05:34 <jmcbride> I have 10 folks registered. 17:05:50 <jmcbride> 1 red hatter has confirmed but not registered 17:05:57 <jmcbride> 2 are TBD in person attendance. 17:06:11 <kiall> Cool - So 10-13 in person? 17:06:31 <jmcbride> 14 if you include me. 17:06:44 <kiall> Cool :) This should be fun! 17:06:55 <betsy> I'm looking forward to it 17:06:56 <mugsie> yup 17:07:05 <jmcbride> yep. I'm hoping to get some tasty austin lunch staples too, so keep your fingers crossed. 17:07:06 <kiall> I also see you've updated the agenda you wrote up 17:07:08 <kiall> #link https://etherpad.openstack.org/p/DesignateAustinWorkshop2014-01 17:07:22 <jmcbride> Yes, I was hoping to get feedback from you guys 17:07:58 <kiall> Looks good, I think on the morning of day 1 we can fill in the "Design sessions" topics, based on "Review Icehouse progress and start blueprint review" 17:08:10 <jmcbride> kiall: agreed 17:08:32 <kiall> #action kiall find 2 hours to write mini-dns spec ahead of time -_- 17:08:57 <kiall> I'm still on the fence over ^ - hence this would be a great time to discuss it ;) 17:08:58 <jmcbride> kiall: that would be great - or at least refine it in your head and we can do a bit of whiteboarding 17:09:12 <jmcbride> kiall: we also have some feedback for you on it 17:09:22 <artom> kiall, actually, about that... you want to discuss that now? 17:09:27 <ekarlso> hey folks 17:09:39 <kiall> myself and mugsie have had a few pub conversations about it, generally refining more each time, but I'm not sure I'm in a great position to articulate it correctly yet 17:10:06 <jmcbride> kiall: I'm adding it to the agenda to specifically address 17:10:08 <kiall> The act of writing it down should help get the final pieces cleared up 17:10:41 <kiall> Cool - Thanks Joe 17:11:29 <jmcbride> "The design sessions priorities" section is where you can add more topics you would like to address outside of icehouse 17:11:30 <kiall> So, other than HP and RAX folks, are there any other folks here today that will be there in person? or hangouts? 17:12:01 <jmcbride> If we have time left over, we can simply start at the top of all known blueprints and prioritize/discusss 17:12:03 <rjrjr_> i'll be on via hangout 17:12:21 <jmcbride> Artom will also be on hangout 17:12:45 <kiall> rjrjr_: Great :) So, we'll have a few via hangouts. 17:13:06 <artom> If you go out for food/beers, I'll bring my laptop to a local Irish pub ;) 17:13:11 <jmcbride> Rackspace will also have our DNS operations lead in person (Jason Bratton). 17:13:15 <ekarlso> I own't be there :p, hangout though ! 17:13:20 <kiall> jmcbride: "If we have time left over," <-- you mean today, or in person? 17:13:31 <jmcbride> ekarlso, can you register so I know to expect you? 17:13:47 <jmcbride> ekarlso: in person during workshop 17:13:59 <jmcbride> ^kiall: in person during workshop 17:14:26 <jmcbride> I'm not sure I want to take you guys to an irish pub in Austin 17:14:35 <jmcbride> Your expectations will be TOO high! 17:14:37 <kiall> jmcbride: perfect :) The last week, myself / mugsie / ekarlso's heads has been been in Ops mode rather than Dev mode ;) 17:15:05 <mugsie> jmcbride: I learnt a long time ago about irish pubs abroad :) 17:15:19 <kiall> Yea - There always .. Interesting places ;) 17:15:34 <artom> Hey, having been in Scotland (not the same, I know) and here in Montreal... We're doing quite OK here :) 17:15:52 <jmcbride> nice, well, I'll try and find an Austin favorite that measure up... 17:16:08 <kiall> rjrjr_ / artom - Whats your TZ difference to Austin? And - Are they any topics you have a particular interest it? 17:16:21 <kiall> (To try and schedule those conversations for your working hours!) 17:16:33 <artom> I'm EST, so +1 17:16:41 <artom> Reverse DNS for me, mostly 17:16:48 <artom> Possible better BIND support 17:16:58 <rjrjr_> MST 17:17:25 <artom> It'll depend on what I get fed in my backlog and what gets decided, if anything 17:18:13 <jmcbride> Sweet, so you guys will be an hour in both directions timezone wise. Will the proposed times be cool? 17:18:19 <kiall> Okay - rjrjr_ - could you add any topics (even if there listed on the page) that you're interested in to the bottom of https://etherpad.openstack.org/p/DesignateAustinWorkshop2014-01 17:18:35 <rjrjr_> yes 17:18:37 <kiall> artom: feel free to update what I put down for you ;) 17:19:03 <kiall> Thanks! 17:19:21 <vinod1> richm: Are there any topics that you would be interested in? 17:19:40 <artom> kiall, sure, I'll keep that info fresh. 17:19:44 <jmcbride> richm - welcome, do you know if you will be able to make it (at least via hangout) 17:20:19 <jmcbride> artom & rjrjr_ - can you send me your google usernames? 17:21:14 <kiall> Let's take that as a yes ;) Otherwise mugsie has them 17:21:19 <rjrjr_> rjrjr@cox,net 17:21:43 <rjrjr_> yes 17:22:16 <jmcbride> One other note about next week priorities, a dev from our Load Balancer team will also be attending (his name is jorge). He has some ideas about "global load balancing". 17:23:05 <kiall> Worth mentioning - next weeks Wednesday IRC meeting will obviously be cancelled, as we'll have just wrapped up the few days in person .. Just making sure that's clear :) 17:23:26 <kiall> jmcbride: excellent :) 17:24:14 <kiall> Anyway - So, Since we have no pre-prepared agenda today, if anyone has anything to bring up, or that they want to discuss, jump in at any time please :) 17:24:51 <artom> Well, re mini-dns... 17:25:01 <artom> Or you want to save that for the workshop? 17:25:08 <kiall> betsy: you're Blacklisted Domains review, is that ready to look over? Last time i looked, you had left a comment saying "not ready yet" :) 17:25:28 <betsy> I've got some changes that I'll submit today and that should be it 17:25:46 <betsy> Sorry, but I've been ill 17:26:02 <kiall> artom: I think we need to get that written up, it's a complex idea with lots of tradeoffs.. And I know I personally still haven't quite got the idea clear in my head 17:26:07 <kiall> betsy: unacceptable ;) 17:26:14 <betsy> :) 17:26:32 <kiall> No worries - I just didn't want it to sit there ready for review for ages without getting looked at ;) 17:27:27 <kiall> Okay - Well, If nobody has anything else, we can call it a day? 17:27:57 <jmcbride> yep, have a great week, see ya'll in Austin! 17:28:05 <kiall> jmcbride: BTW 17:28:11 <eankutse> see you 17:28:17 <artom> kiall, well, I just have... an "alternative"? 17:28:29 <artom> Not comments about the idea itself. 17:28:55 <kiall> jmcbride: Myself and mugsie be in Austin on Sunday, and staying in the Radission .. Just FYI 17:29:02 <kiall> artom: an alternative? 17:29:43 <artom> Well, instead of implementing mini-dns, the same effect can be achieved by mandating the multi backend with a PowerDNS master. 17:29:47 <jmcbride> kiall: sweet, I emailed all attendees my contact info. I might be able to slip away for dinner/drinks Sunday, so let me know. 17:30:17 <kiall> jmcbride: depends how tried we are ;) Land at like 6am or something. 17:31:06 <kiall> artom: That's certainly an option, but has some drawbacks in terms of the potential phase 2 features.. 17:31:45 <kiall> Inbound AXFR, RFC / nsupdate Dynamic DNS, anything that modifies the content of the zone, would be really really awkward to fit into that model 17:32:29 <artom> Do we need to allow zone modification by means other than the API? 17:33:15 <kiall> I think we do, and I know rjrjr_ would agree ;) one of the pieces he discussed in the past was accepting RFC / nsupdate dynamic DNS updates 17:33:20 <eankutse> artom: we should not rule out use of nsupdate 17:33:38 <mugsie> afik redhat openshift needs it as well 17:33:42 <jmcbride> kiall/artom: regarding mini-DNS, I think it sounds feasible, but I think it will only work for use cases in which either zone transfers or dynamic dns protocol is supported. 17:33:49 <kiall> RedHat I'm sure are interested in that too - OpenShift / Origin requires a a RFC style dynamic DNS server 17:34:22 <artom> And so mini-dns would implement said RFC-style server? 17:34:42 <artom> And reflect any changes it receives via not-API in Designate's state? 17:34:45 <kiall> jmcbride: absolutely - this would never be a target of end user queries - that would be an epic failure waiting to happen IMHO ;) 17:35:39 <kiall> artom: yea, so, RFC dynamic DNS is actually quite simple to impelement (with the exception of transactional / atomic commits of whole changeset - but that's another topic to solve independently) 17:35:43 <jmcbride> kiall: agreed. My comment was targetted at how to get zone data transferred from master to slaves. 17:36:00 <kiall> When I say simple - I mean, simple once you have the other pieces in place 17:36:59 <jmcbride> kiall: I think we will benefit from that 2 hour homework assignment you gave yourself… 17:37:20 <kiall> jmcbride: and lots of RFC reading ;) 17:37:38 <jmcbride> kiall: before bed is the best time for that 17:37:55 <kiall> You'll never get past the first sentence that way 17:38:34 <rjrjr_> our mini-dns also needs to respond to refresh requests from slaves 17:38:48 <kiall> artom: Anyway - I know I see huge advantages by being able to alter the data we have via the standard DNS mechanisms, I think supporting that is key if we go with the "mini-dns" suggestion , or the enforce powerdns alternative route you suggested 17:39:18 <eankutse> Kiall: regarding RFC's that you plan to read 17:39:31 <eankutse> can you tell us which ones you plan to look at? 17:39:34 <kiall> eankutse: I meant you guys ;) I've read them all 17:39:37 <eankutse> I am interested as well 17:39:42 <eankutse> :-) 17:39:57 <eankutse> sleep reading RFC's? 17:40:19 <kiall> rfc1034, rfc1035, rfc2181 are some the main ones 17:40:25 <eankutse> cool 17:40:28 <artom> kiall, was just throwing the alternative out there. It seemed like mini-dns was reimplementing existing stuff for benefits that were unclear to me. But if we want storage <-> backend "bidirection" communications, I don't think we have the choice... 17:40:41 <kiall> 2136 is nsupdate / RFC style dynamic DNS 17:41:10 <artom> Unless we start doing funky stuff like polling the PowerDNS db to import changes into storage... 17:41:14 <kiall> Ehh - And I'm sure theres 100 others that aren't coming to mind right now 17:41:49 <eankutse> That's a good start 17:41:59 <eankutse> there'll be references from those to others 17:42:01 <kiall> artom: Well, powerDNS supports custom backends, lua based from memory, but we enter pretty dangerous territory there with ever becoming core OpenStack ;) 17:42:33 <artom> That's a whole other level of meta right there... 17:42:47 <kiall> And - Choosing PowerDNS as a core piece, as much as I like that idea, would likely be a roadblock for us over time as people have their favourite 17:42:49 <artom> "We put a backend in your backend so you can DNS while you DNS" ;) 17:43:11 <kiall> artom: hah :) Yep 17:43:36 <jmcbride> We should have a blueprint for the idea of supporting bidirectional updates (e.g. using dynamic DNS to update zones). 17:44:28 <kiall> oh eankutse .. rfc2845 would be another core piece.. providing a method of identification of users making queries etc to this 17:45:07 <jmcbride> http://i.imgur.com/kQ7DHB4.jpg 17:45:22 <kiall> lol 17:45:36 <artom> jmcbride, I didn't think anyone would actually do that :D 17:45:42 <artom> Bravo. 17:46:01 <kiall> artom: CloudFlare was (is?) based on PowerDNS and a custom backend for it ;) 17:46:10 <kiall> Okay - Well, I think it's time to wrap this up and call it a day :) 17:46:12 <eankutse> Kiall: thx. Keep them coming :-) 17:46:20 <kiall> hub_cap got here, so .. everyone leave please. 17:46:23 <kiall> ;) 17:46:28 <hub_cap> YEA BOI 17:46:48 <hub_cap> hi kiall ;) i totally saw a half past / half 30 thing on tv the other day 17:46:49 <artom> Hey, we got 15 more minutes officially, I intend to fill every last second with more memes :D 17:47:01 <kiall> Okay - Thanks all :) Cya next week ! 17:47:12 <betsy> cya 17:47:18 <kiall> #endmeeting