17:00:04 <Kiall> #startmeeting Designate
17:00:08 <openstack> Meeting started Wed Mar 18 17:00:04 2015 UTC and is due to finish in 60 minutes.  The chair is Kiall. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:00:09 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
17:00:12 <openstack> The meeting name has been set to 'designate'
17:00:14 <Kiall> Hey folks, who's about?
17:00:23 <rjrjr_> o/
17:00:26 <vinod1> o/
17:00:35 <timsim> o?
17:00:47 <elarson> o/
17:01:03 <Kiall> So, considering k3 tomorrow, I'm mixing up the regurarily scheduled agenda.. Todays agenda is:
17:01:06 <Kiall> Kilo Release Status, More Kilo Release Status, More More Kilo Release Status. ;)
17:01:31 <Kiall> Any objections?
17:01:41 <vinod1> none
17:01:47 <elarson> +1
17:01:52 <Kiall> #topic Kilo Release Status
17:02:07 <Kiall> #link https://launchpad.net/designate/+milestone/kilo-3
17:02:29 <Kiall> Looking better than last week :) But, still lots up for review and to complete.
17:02:52 <Kiall> #link https://review.openstack.org/#/q/status:open+(project:openstack/designate+OR+project:openstack/python-designateclient+OR+project:openstack/designate-specs),n,z
17:03:53 <Kiall> So, Graham's been working on the validation cleanup (i.e. getting validation into V2 that actually works).. That's the only BP left on the k3 page, but we're missing one.. Pools->DB as discussed at the mid-cycle.
17:04:12 <Kiall> Besides those, are we missing anything?
17:04:50 <vinod1> nothing else comes to mind
17:05:11 <Kiall> Cool, lets start with those validation ones then.. mugsie - can you give a quick status?
17:05:14 <timsim> Yeah I think that's it
17:05:25 <mugsie> https://review.openstack.org/#/c/163131/
17:05:25 <mugsie> https://review.openstack.org/#/c/164213/
17:05:25 <mugsie> https://review.openstack.org/#/c/164241/
17:05:25 <mugsie> https://review.openstack.org/#/c/164326/
17:05:25 <mugsie> https://review.openstack.org/#/c/165070/
17:05:27 <mugsie> https://review.openstack.org/#/c/165453/
17:05:30 <mugsie> https://review.openstack.org/#/c/165453/
17:05:33 <mugsie> are the review that are good to go
17:05:57 <mugsie> my feature branch is so long right now that it is getting difficult to maintain it
17:06:22 <vinod1> thanks mugsie - the list helps to start reviewing
17:06:45 <mugsie> there is some random commits in there -  but they are the chain right now, so merging them would be helpful
17:07:31 <Kiall> mugsie: https://review.openstack.org/#/c/165453/ failing on pep8,py27 etc BTW
17:07:39 <timsim> Does the API change for all those objects need to go in k3, or can that be rc1?
17:07:42 <mugsie> yeap, i saw that just before the meet started
17:07:52 <mugsie> timsim: I would like it to be k3 personally
17:07:59 <mugsie> at least the start
17:08:13 <mugsie> v1 may be pushed, but v2 not so much
17:08:35 <Kiall> timsim: ideally we try and land as much in k3 as possible, techincally feature freeze after that, so if there's bug fixing is no issue.. That said, we can allow new stuff in where we deem it important ;)
17:09:04 <timsim> Cool, it's not like it's some big functionality breaking feature anyway
17:09:35 <Kiall> that chain leads up to https://review.openstack.org/#/c/164327/, which leads to finally having validation fixed between both APIs - and validation of record data etc etc inn V2
17:09:53 <Kiall> I believe those are the final bits, right mugsie?
17:09:58 <mugsie> nope
17:10:06 <mugsie> there is one more on my workstation
17:10:16 <mugsie> that needs all ^ to go through :)
17:10:16 <Kiall> Yea, that's the seconds "leads to" ;)
17:10:41 <mugsie> ah, yeah
17:11:14 <Kiall> Ideally, we can spend the second half of the hour looking over the reviews, so we can get any Q+A done while everyone is at a keyboard :)
17:12:07 <vinod1> Kiall: Did you get a chance to look at https://bugs.launchpad.net/designate/+bug/1433585?
17:12:08 <openstack> Launchpad bug 1433585 in Designate "Zone creates in Tempest occasionally go to ERROR" [High,Triaged]
17:12:10 <Kiall> Okay - Let's move on to the Pools DB (we'll come back to this..) Part 1/3 of that change is up - https://review.openstack.org/#/c/165146/ - that's the refactor of the existing code that changes as part of it.
17:12:33 <Kiall> vinod1: I started digging in, but am not quite sure what the trigger is yet
17:13:34 <Kiall> Part 2 of the Pools -> DB change will be up in the next 2-3 hours, it creates the vaious objects, and builds them up based on the config file - it has all the bits discussed at the mid-cycle, like the "targets" split etc
17:13:41 <Kiall> #link https://etherpad.openstack.org/p/designate-mid-cycle-jan-2015
17:14:10 <Kiall> But, Part #3/3 - Actually storing it in the DB will have to come after k3 :(
17:14:51 <ekarlso> o/ btw
17:14:54 <ekarlso> i'm back and forth
17:15:09 <Kiall> Right now, the config looks like this: http://paste.openstack.org/show/193189/ - which turns into this: http://paste.openstack.org/show/193191/
17:15:13 <Kiall> once parsed
17:15:45 <Kiall> (note that JSON-looking blob is a to_dict of the object, rather than 1:1 what the API looks like
17:16:10 <Kiall> getting this in unblocks timsim's agent stuff, dynect, and pushing the Akamai driver..
17:16:18 <Kiall> pushing the akamai driver we have up*
17:16:39 <Kiall> Any thoughts/comments on ^?
17:16:56 <timsim> Can't wait to have it.
17:17:07 <Kiall> Yea, it's been a long time coming.. I apologize ;)
17:17:26 <Kiall> I will have the config version up + ready before I leave the building tho.. and I'm in the office ;)
17:17:59 <Kiall> the final changeset, actually storing it in the DB etc will come after - but I decided to take this route to unblock other stuff as quick as I could.
17:18:54 <mugsie> just need to get it in :D
17:19:19 <mugsie> but we needed to unblock agent, which unblocks old style backends
17:19:29 <mugsie> which is needed for K-final
17:19:35 <Kiall> Yep - Exactly..
17:20:09 <Kiall> So - I'd love to spend some time now giving people a chance to look through the reviews there, get any Q+A etc talked through while were all here :) Thoughts?
17:20:16 <timsim> That sounds good
17:21:05 <Kiall> #topic Final K3 Push - Review/Q+A/Mergefest (Hopefully? ;))
17:21:21 * Kiall shuts up and gives people a few mins to start looking though changes.
17:25:12 <vinod1> mugsie: In https://review.openstack.org/#/c/163131/4/designate/objects/zone_transfer_request.py, should there be a format specifier for domain_name as domainname (line 62)
17:25:53 <mugsie> vinod1: it doesnt matter, as that is loaded from the DB, and the user is never allowed set it
17:25:55 <Kiall> vinod1: agh, I thought I found all the differences a few patchsets back :(
17:26:05 <mugsie> (its read only)
17:26:21 <mugsie> so we shouldn't be valdating it (that value never gets stored)
17:28:34 <timsim> I didn't have anything else on that PS
17:33:53 <mugsie> I just realised that I had a lot of code there
17:34:07 <Kiall> ekarlso: on https://review.openstack.org/#/c/165514/, I know why we had reset_notifications now ;) I've left a comment with details.
17:35:03 <vinod1> mugsie: Re https://review.openstack.org/#/c/164213/, it looks like all the extensions have moved to admin extensions. So is there a need for the config setting enabled_extensions_v2?
17:35:11 <Kiall> mugsie: yea, I think I forgot to say "quick review now rather than detailated examination etc ;)" - more for making sure any design Q's etc are worked out
17:35:46 <mugsie> vinod1: we should be able add extentions to v2 in the future
17:35:54 <mugsie> or allow users to add them
17:36:17 <Kiall> Yea, users can add extensions out-of-tree, so keeping that makes sense IMO..
17:36:38 <vinod1> good to know that my nearly abandoned admin review is of some use
17:37:24 <mugsie> vinod1: yeah, i did a C+P job on it ;)
17:37:35 <timsim> +A'd admin api
17:37:43 <mugsie> :)
17:40:37 <Kiall> Also - There's another issue in the gate right now, exposed by paul's tempest stuff - https://review.openstack.org/#/c/165390/
17:40:57 <Kiall> Though, I've not checked why the pdns gate has failed 3 times in a row on that one..
17:42:57 <timsim> mugsie: https://review.openstack.org/#/c/164241/5 doesn't do anything atm right? It's waiting on the errors future stuff will raise?
17:43:05 <mugsie> timsim: yup
17:46:33 <timsim> +A'd validationmiddleware
17:46:38 <mugsie> :)
17:48:54 <Kiall> Okay - With 10 (well.. 12) mins left, Let's move to Open Discussion in case anyone has something besides K3 stuff ;)
17:48:55 <Kiall> #topic Open Discussion
17:49:06 <Kiall> So - Anyone have something besides K3?
17:49:34 * Kiall gives people a few mins to get their heads out of Gerrit ;)
17:50:02 <timsim> I don't have anything :)
17:50:21 <mugsie> I have nothing :)
17:51:03 <Kiall> I think mugsie's just estatic a few of his reviews won't need more rebasing :P
17:51:15 <mugsie> yup
17:51:30 <elarson> if folks wanted to take a peak at the hook api review
17:51:33 * elarson is grabbing a link
17:52:32 <elarson> https://review.openstack.org/#/c/164748/
17:52:39 <Kiall> and https://review.openstack.org/#/c/164747/ :)
17:52:50 <elarson> no rush or anything, but just wanted to throw it out there
17:52:57 <elarson> indeed!
17:53:09 * timsim is excited to try that out
17:53:19 <mugsie> its on my list, so post K3 100% will look
17:53:55 <vinod1> one comment that I had - what will happen to the domain when the hook blocks a domain create?
17:54:56 <elarson> vinod1: good question. since the hook has access to the designate code, it can do the right thing regarding setting it in the error state.
17:55:04 <vinod1> we also need to ensure that the recovery/sync threads do not go and create the domain later on
17:55:34 <elarson> right, the hook is responsible for being a good designate citizen in that regard and making sure things are in a reasonable state
17:55:39 <Kiall> elarson: One pattern I've liked is the base WSGI mdidleware patten, https://github.com/openstack/oslo.middleware/blob/master/oslo_middleware/base.py#L21
17:56:01 <Kiall> your before ~= their process_request
17:56:39 <Kiall> where, if process_request returns a value != None, it never even calls the wrapped method, and treats the return value as the result
17:57:06 <elarson> yeah, I'm not totally sold on my before/after pattern, so I can imagine that could change based on what the requirements of an org needed to accomplish
17:57:11 <Kiall> I wonder if that kinda patten fits, rather than before *needing* to raise an exception to prevent the hooked code from  running
17:57:23 <timsim> One thing I saw yesterday. If you're at database migration 50. And you do designate-manage upgrade 45, it downgrades to 45. Same thing happens the other way :P
17:57:41 * elarson has to head out for an appt
17:57:43 <elarson> thanks all!
17:57:46 <mugsie> o/
17:57:59 <Kiall> Okay, I'm going to call the meet here.. If people are still checking those K3 items, let's move to #openstack-dns :)
17:57:59 <Kiall> and - don't forget this one https://review.openstack.org/#/c/165390/ ;)
17:58:19 <Kiall> Thanks all :)
17:58:24 <Kiall> #endmeeting