14:00:08 <acabot> #startmeeting watcher 14:00:09 <openstack> Meeting started Wed Jan 20 14:00:08 2016 UTC and is due to finish in 60 minutes. The chair is acabot. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:10 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:13 <openstack> The meeting name has been set to 'watcher' 14:00:15 <vincentfrancoise> o/ 14:00:16 <acabot> hi 14:00:20 <brunograz> hi 14:00:23 <dtardivel> o/ 14:00:23 <edleafe> \o 14:00:34 <jed56> -o/ 14:00:51 <slaweq> hello 14:01:04 <jwcroppe_> o/ 14:01:18 <acabot> our agenda for today https://wiki.openstack.org/wiki/Watcher_Meeting_Agenda#01.2F20.2F2016 14:01:25 <tkaczynski> hi 14:01:36 <vmahe> hi 14:02:16 <acabot> #topic Announcements 14:02:27 <tpeoples> \\o 14:02:33 <acabot> quick update regarding mitaka-2 14:02:54 <acabot> #info : 11 BPs en 41 bugs targeted for mitaka-2 14:03:00 <sballe> o/ 14:03:17 <acabot> it is more BPs than Nova project :-) 14:03:24 <acabot> so thanks to all of you 14:03:31 <jwcroppe> +10000 14:03:39 <sballe> one quick annoucement from Intel. bzhou has decided to leave intel and will be replaced with edwin ... 14:03:59 <acabot> just to remind you that we had only 3 BPs for mitaka-1 ! 14:04:04 <sballe> I cannot remember edwin;s irc 14:04:14 <gzhai> I'm edwin 14:04:24 <acabot> hella gzhai 14:04:31 <gzhai> hello 14:04:31 <sballe> Welcome gzhai :-) 14:04:34 <acabot> welcome 14:04:43 <vtech> o/ 14:04:45 <jed56> Welcome gzhai! 14:04:47 <tpeoples> hi gzhai 14:04:57 <vincentfrancoise> welcome gzhai :) 14:05:03 <gzhai> hello all:) 14:05:18 <sballe> acabot: I am working with nishi and team to get the bps in asap. she has been super busy reworkign the story around the demo and showing it to senior folks at intel 14:05:19 <acabot> gzhai : could you please register to watcher drivers ? https://launchpad.net/~watcher-drivers 14:05:36 <gzhai> acabot: ok 14:05:47 <acabot> sballe: is gzhai the new Intel core contrib ? 14:05:56 <sballe> yes please 14:06:05 <sballe> acabot: yes please 14:06:35 <acabot> #action dtardivel update gzhai & bzhou core dev status on launchpad 14:06:47 <sballe> bzhou told me he still want to work on watcher when he moves over to his new company but I think let's see if they let him and maybe he can earn his core status the regular way 14:06:54 <acabot> I suppose we remove the core feature to bzhou ? 14:07:08 <acabot> sballe : +1 14:07:17 <tkaczynski> hi Tomasz here, should I also register to watcher drivers? 14:07:29 <acabot> #action acabot reassign BPs & bugs to gzhai 14:07:32 <sballe> tkaczynski: Welcome too 14:07:52 <sballe> yes please register to watcher drivers tkaczynski 14:08:09 <acabot> tkacynski : hi Tomasz, yes you can 14:08:34 <jed56> everybody can register to watcher drivers for contributing : -) 14:08:35 <acabot> #action gzhai tkaczynski register to watcher-drivers 14:08:46 <acabot> any other annoucement so far ? 14:09:04 <sballe> no from me 14:09:14 <acabot> #topic Review Action Items 14:09:17 <jwcroppe> just that it will be nice to see everyone in Austin in two weeks :) 14:09:47 <acabot> jwcroppe : :-D, we will be 10 in Austin which is great ! 14:09:50 <sballe> +1' 14:09:50 <dtardivel> tkaczynski: what is your username on Launchpad ? 14:11:10 <jwcroppe> btw, for mid-cycle - we will order lunch in every day 14:11:18 <sballe> +10000 14:11:19 <tkaczynski> dtardivel I don't have username there yet 14:11:22 <acabot> we need a status on https://blueprints.launchpad.net/watcher/+spec/optimization-threshold, gzhai do you plan to take it ? 14:11:24 <sballe> so working lunches ;-) 14:11:42 <acabot> sballe : ok ;-) 14:11:58 <gzhai> acabot: I assume that 14:12:03 <tkaczynski> I'm just starting to organize myself here, looking around :) 14:12:13 <sballe> is bzhou in the meeting? 14:12:19 <acabot> gzhai : ok thx 14:12:42 <gzhai> sballe: seems no 14:12:58 <sballe> gzhai: ok I was just checking 14:13:14 <sballe> so gzhai thx for taking that over 14:13:27 <acabot> #info the spec regarding efficacy indicator is in review #link https://review.openstack.org/#/c/266446/ but I need to add work items 14:14:42 <seanmurphy> i have a basic q which i think touches on this 14:14:45 <jwcroppe> acabot: thanks, any other urgent specs to review? 14:15:04 <seanmurphy> do we assume that there can be multiple strategies which can achieve a specified goal then? 14:15:18 <seanmurphy> and the efficacy indicator is the way of arbitrating between them 14:15:19 <seanmurphy> ? 14:15:20 <acabot> jwcroppe : no 14:15:47 <acabot> seanmurphy : I think we should discuss that in specs 14:15:49 <jwcroppe> I think we decided that we'd start with 1 strategy<-->1 goal to start with ? 14:15:56 <seanmurphy> ok 14:16:03 <seanmurphy> that seems more reasonable to me 14:16:07 <acabot> jwcroppe: right 14:16:26 <acabot> #info a lot of improvement on Watcher doc is ongoing with many reviews 14:16:51 <acabot> jwcroppe: I'd like to have your review on https://review.openstack.org/#/c/266461/ 14:17:17 <acabot> vmahe did a great job putting sequence diagrams in Watcher doc 14:17:36 <sballe> I need to leave. Sorry about that 14:17:47 <acabot> and I think its important for everyone who wants to understand how Watcher is working right now 14:17:57 <acabot> sballe: ok 14:18:03 <gzhai> sballe: bye 14:18:15 <acabot> #action jwcroppe review https://review.openstack.org/#/c/266461/ 14:18:32 <sballe> bye 14:18:34 <jwcroppe> acabot: will do 14:19:02 <acabot> #info dynamic loading of Actions handlers in the Watcher Applier feature is now available in Watcher 14:19:17 <jwcroppe> that's excellent 14:19:32 <acabot> but there is still a bit work to do 14:19:42 <acabot> on contract and oslo 14:19:49 <jwcroppe> acabot: in mid-cycle can we carve out an hour time slot (maybe one evening Mon/Tue) to see demo of some of these new feature? 14:19:50 <seanmurphy> is there a list of known actions? 14:20:25 <acabot> jwcroppe : yes I will add on hour of demo time on tuesday 14:20:41 <jwcroppe> ok, great - we can make it maybe last topic in evening 14:20:48 <jed56> You can take a look to the entry points 14:20:57 <jed56> but we can update the documentation to list them 14:21:16 <jed56> we have ( Migrate,Nop, Sleep, Change hypervisor State ) 14:21:27 <seanmurphy> jed56: perhaps i can discuss this offline with you - thanks 14:21:37 <jed56> https://github.com/openstack/watcher/blob/master/setup.cfg#L49 14:21:53 <acabot> #info vincentfrancoise has set up Tempest in Watcher 14:22:07 <jed56> seanmurphy : yes of course 14:22:12 <acabot> and he needs reviews on https://review.openstack.org/#/q/status:open+project:openstack/watcher+branch:master+topic:bp/tempest-basic-set-up 14:22:42 <acabot> anyone want to review tempest ? 14:22:53 <tpeoples> yeah i will when i get time 14:23:08 <acabot> #action tpeoples review Tempest set up https://review.openstack.org/#/q/status:open+project:openstack/watcher+branch:master+topic:bp/tempest-basic-set-up 14:23:19 <acabot> tpeoples: thx 14:23:41 <acabot> #info tpeoples is working on external API versionning WIP 14:24:00 <tpeoples> yeah, hoping to get that done in the next 1-2 days 14:24:12 <tpeoples> i decided to follow what magnum is doing with their clients 14:24:20 <tpeoples> think that will get us the flexibility we need 14:24:32 <acabot> tpeoples: it would be great to have it for mitaka-2 14:24:42 <jwcroppe> tpeoples: can you put me on the review for that when you post it up? 14:24:43 <tpeoples> yep, will try :) 14:25:41 <acabot> #info watcher-client has been updated to reflect changes in Watcher 14:25:55 <jed56> tpeoples : what are the clients you want to support ? 14:26:15 <jed56> neutron, keystone, ... ? 14:26:16 <tpeoples> jed56: just going to do nova, ceilometer, glance, cinder for now 14:26:25 <tpeoples> but it will be easy to add more 14:26:30 <tpeoples> +keystone 14:26:30 <jed56> okay cool 14:26:53 <acabot> there is still on review open on watcher-client #link https://review.openstack.org/#/c/266473/ 14:27:37 <acabot> seanmurphy : do you want to share your work in progress now ? 14:27:52 <seanmurphy> we have not had much to update tbh 14:27:52 <gzhai> tpeoples: means watcher-client can call nova/ceilometer's API? 14:27:58 <seanmurphy> we said we would produce a spec for today 14:28:11 <seanmurphy> on how to implement our load consolidation mechanism in watcher 14:28:17 <seanmurphy> we will produce a draft of it later 14:28:32 <tpeoples> gzhai: implicitly yes, but you won't be able to do `watcher ceilometer-meter-list` or anything like that 14:28:41 <seanmurphy> i sent a link of the draft gdoc to you so you can see 14:28:43 <acabot> seanmurphy : seems great, thx for the update 14:28:47 <tpeoples> just that there will be a standard way to fetch other project's clients from the watcher backend 14:28:54 <jed56> seanmurphy : you can follow the doc ( https://github.com/openstack/watcher/blob/master/doc/source/dev/plugins.rst) 14:28:54 <seanmurphy> tbh it’s our first time writing such a spec 14:29:05 <seanmurphy> so we’re not sure if we’re hitting all the right notes 14:29:10 <gzhai> tpeoples: got it. 14:29:17 <seanmurphy> jed56: thanks - will do 14:29:38 <seanmurphy> the process is that we git push to watcher-specs, right? 14:30:13 <acabot> seanmurphy : no :-p 14:30:41 <edleafe> tpeoples: is the watcher API stable enough yet to consider controlled versioning, like microversions? 14:30:45 <acabot> #action seanmurphy look at https://wiki.openstack.org/wiki/Watcher/Contributing#Your_first_commit 14:30:54 <seanmurphy> ok - will do 14:31:14 <tpeoples> i would say not yet edleafe 14:31:19 <acabot> seanmurphy : we will iterate on your spec through Gerrit 14:31:26 <seanmurphy> ok 14:31:33 <edleafe> tpeoples: ok, I didn't think so. 14:31:48 <acabot> seanmurphy : please push it for reviews in watcher-specs/specs/mitaka/approved folder 14:32:02 <jwcroppe> edleafe: IIRC, when we discussed that in Tokyo, we decided to get through the POCs and determine which APIs, etc. are really needed, which need to be updated, etc. and do some 'clean up' before we get to the full-blown microversions, etc. that folks will expect once 'big tent' 14:32:02 <seanmurphy> right - and it should follow the format in https://github.com/openstack/watcher-specs/blob/master/specs/mitaka-template.rst 14:32:36 <acabot> seanmurphy : yes 14:32:38 <edleafe> jwcroppe: that sounds sane 14:32:49 <acabot> #topic Blueprint/Bug Review and Discussion 14:33:24 <acabot> I'd like to have a quick bug list review for mitaka-2 as we still have 17 bugs opened 14:33:33 <acabot> #link https://launchpad.net/watcher/+milestone/mitaka-2 14:34:01 <tkaczynski> dtardivel: I added myself to watcher drivers, waiting for approval 14:34:38 <jwcroppe> acabot: need to attend another meeting, will review any new action items you post for me... I will also review the spec later this afternoon referenced earlier 14:34:48 <dtardivel> tkaczynski: done 14:35:21 <acabot> we need folks for #link https://bugs.launchpad.net/watcher/+bug/1536191 #link https://bugs.launchpad.net/watcher/+bug/1525830 #link https://bugs.launchpad.net/watcher/+bug/1523513 14:35:24 <openstack> Launchpad bug 1536191 in watcher "Bad place for checking goal existence when creating a new audit template" [Low,New] 14:35:26 <openstack> Launchpad bug 1525830 in watcher " Refactored Watcher codebase to add py34 support" [Low,Confirmed] 14:35:27 <openstack> Launchpad bug 1523513 in watcher "Add tests to primitive" [Medium,In progress] 14:36:13 <acabot> and https://bugs.launchpad.net/watcher/+bug/1521636 14:36:14 <openstack> Launchpad bug 1521636 in watcher "MessagingCore - reference to a self.client only defined in children" [Low,In progress] 14:37:43 <acabot> tpeoples: you want to talk about https://bugs.launchpad.net/watcher/+bug/1510188 ? 14:37:44 <openstack> Launchpad bug 1510188 in watcher "Creation of a watcher audit with an unknown audit template generates bad error" [Low,Confirmed] - Assigned to Taylor Peoples (tpeoples) 14:38:07 <tpeoples> yeah, so the fix for this is a tad more complicated i think 14:38:37 <tpeoples> a user could use an audit template to create an audit, then update the audit template to use a totally other goal 14:38:44 <tpeoples> and the audit has no trail of that change 14:38:56 <tpeoples> so the history of the audit would make no sense 14:39:38 <tpeoples> likewise, if someone deletes the audit template, the audit has no way to "know" what goal it had originally 14:40:06 <vincentfrancoise> tpeoples: It's kind of linked to #link https://bugs.launchpad.net/watcher/+bug/1532843 then 14:40:07 <openstack> Launchpad bug 1532843 in watcher "Can create Audit with invalid parameters" [Low,Confirmed] 14:40:21 <tpeoples> i propose that the audit template is simply a holder of the metadata for the audits, and when an audit is created it copies over that metadata so there is no hard link there 14:40:27 <acabot> tpeoples: true but is it the problem targeted by this bug ? 14:40:30 <tpeoples> but i think that is going to require a BP 14:41:01 <acabot> tpeoples: right we need a BP & specs 14:41:01 <tpeoples> yes acabot, because if i fix it, it has implications for listing audits when doing the validation of the audit template 14:41:21 <jed56> +1 14:41:35 <acabot> tpeoples : could you submit a BP wih your first assumption ? 14:41:47 <tpeoples> yes. i will outline the problem in more words 14:42:11 <tpeoples> but long story short, i'm not going to get that bug done since i think we need a more complete solution 14:42:32 <acabot> #action tpeoples submit a BP regarding audit template history management 14:42:49 <acabot> tpeoples : ok I will schedule it for mitaka-3 14:42:55 <tpeoples> sounds good acabot 14:43:16 <acabot> #action acabot schedule https://bugs.launchpad.net/watcher/+bug/1532843 & https://bugs.launchpad.net/watcher/+bug/1510188 for mitaka-3 14:43:18 <openstack> Launchpad bug 1532843 in watcher "Can create Audit with invalid parameters" [Low,Confirmed] 14:43:19 <openstack> Launchpad bug 1510188 in watcher "Creation of a watcher audit with an unknown audit template generates bad error" [Low,Confirmed] - Assigned to Taylor Peoples (tpeoples) 14:43:37 <acabot> #topic Open Discussion 14:43:41 <dtardivel> tpeoples: acabot not sure. bug is only bad return error 14:44:06 <dtardivel> when we try to create an audit with a bad audit template name. 14:44:12 <acabot> dtardivel: you want to fix it with the good error code ? 14:44:45 <tpeoples> the fix for that would break the listing audits (if there are any audits with deleted audit templates) 14:45:08 <dtardivel> I think we can try to return the correct error status in that case. and create a new BP about how to store data for history 14:45:29 <dtardivel> acabot: yes 14:45:43 <dtardivel> acabot: this is the scope of the bug 14:46:19 <tpeoples> dtardivel: let's discuss in irc after the meeting 14:46:26 <acabot> tpeoples: ok 14:46:38 <acabot> cross-project liaison 14:46:58 <acabot> I think it would be great to start having a cross project liaison for Watcher 14:47:19 <acabot> the weekly meeting is at 9pm UTC 14:47:46 <acabot> I'm sure I will be able to atend at 10pm my time 14:47:50 <tpeoples> acabot: there was a note on the ML about this. it looks like the PTL's were encourage to attend the meeting, but if you aren't interested / can't make the meeting then we should find someone else to do it 14:47:55 <acabot> anyone would be interested in doing it ? 14:48:11 <tpeoples> jwcroppe or sballe may be interested in that role? 14:48:34 <acabot> I suppose but none of them is here :-D 14:49:20 <acabot> #action jwcroppe sballe would you be interested in being cross-project liaison for Watcher ? 14:49:34 <acabot> they will discover it ;-) 14:49:45 <tpeoples> :P 14:50:42 <acabot> slaweq: how is it going with watcher ? 14:51:19 <tpeoples> acabot: what's it? 14:51:58 <acabot> slaweq is one of our first Watcher user ;-) 14:52:12 <acabot> but I'm not sure he is still with us 14:52:14 <jed56> (OVH) 14:52:50 <tpeoples> wow i totally misread your last comment, too early :D 14:53:37 <acabot> I just relealized I forgot to mention taskflow integration by jed56 14:54:22 <acabot> #info taskflow integration is now in review #link https://review.openstack.org/#/c/264142/ 14:54:22 <tpeoples> thanks for working on that jed56 14:55:05 <acabot> as we plan to release mitaka-2 on friday EOD, we will need to merge a lot of code tomorrow 14:55:38 <acabot> so please try to review most of the code before 14:56:08 <tpeoples> +1 acabot, lots to do today 14:56:10 <acabot> and set it to -1 if you dont want something to be merged in mitaka-2 14:57:09 <acabot> #info 21 reviews open waiting for mitaka-2 14:58:37 <acabot> #info mid-cycle agenda is available on #link https://wiki.openstack.org/wiki/Watcher_mitaka_mid-cycle_meetup_agenda#Agenda 14:59:06 <acabot> even if you dont plan to be there, please add subjects you would like to be discussed 14:59:23 <acabot> its time to close the meeting 14:59:26 <acabot> thank you 14:59:30 <acabot> bye 14:59:32 <tpeoples> talk to you later 14:59:38 <tkaczynski> bye 14:59:40 <vincentfrancoise> bye 14:59:46 <jed56> bye 14:59:51 <acabot> #endmeeting