14:00:16 <acabot> #startmeeting watcher 14:00:20 <openstack> Meeting started Wed Aug 24 14:00:16 2016 UTC and is due to finish in 60 minutes. The chair is acabot. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:21 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:23 <openstack> The meeting name has been set to 'watcher' 14:00:25 <acabot> hello 14:00:28 <sballe_> hi 14:00:41 <tkaczynski_> hi 14:00:49 <acabot> agenda for today #link https://wiki.openstack.org/wiki/Watcher_Meeting_Agenda#08.2F24.2F2016 14:00:52 <diga> Hi 14:00:57 <tri2sing> hi 14:01:09 <edleafe> \o 14:01:39 <gzhai> hi 14:01:52 <acabot> #topic Announcements 14:02:07 <acabot> #info please vote for promoting hvprash as core on watcher-specs 14:02:14 <acabot> #link http://lists.openstack.org/pipermail/openstack-dev/2016-August/101750.html 14:02:29 <acabot> sballe_ : I'd really like to have your +1 before next week 14:02:44 <sballe_> acabot: was looking at it right now. 14:02:46 <vincentfrancoise> o/ 14:03:18 <acabot> #info welcome to our new Walmart contributors (tri2sing today) 14:03:28 <tri2sing> thanks 14:03:43 <gzhai> welcome 14:03:59 <acabot> tri2sing : I saw you started reviewing specs, thats great, feel free to ask if there is anything you would like to do in the project 14:03:59 <vincentfrancoise> welcome! 14:04:19 <diga> welcome! 14:04:24 <sballe_> +1 14:04:29 <tri2sing> Will certainly do. I though I'd start small. :) 14:04:55 <acabot> tri2sing : ok did you already set up Watcher in devstack ? 14:05:23 <tri2sing> Not yet. I am going through the manual set up first. 14:05:36 <acabot> ok 14:05:46 <acabot> #info notifications for acceptance (or not) of our talks for Barcelona OpenStack Summit will be sent tomorrow 14:06:03 <acabot> finger crossed ;-) 14:06:06 <sballe_> acabot: what was the quality of the talks you reviwed? 14:06:28 <sballe_> I am assuming you reviewed the talks for your track 14:06:31 <diga> do you have the talk link ? 14:06:39 <acabot> sballe_ : I was on the cloud economics tracks so few talks submitted but quite interesting 14:06:58 <sballe_> that sounds like a great subject/track 14:07:20 <acabot> all our talks had a mark above 2 so it was quite well, but I dont know if they have been selected ;-) 14:08:03 <sballe_> that is great news in any case 14:08:06 <acabot> diga : https://etherpad.openstack.org/p/Watcher_abstracts_barcelona2016 14:08:17 <diga> acabot: thx 14:08:23 <acabot> #info Space requested for the design summit is 3 workroom sessions (40 min slots), 1 fishbowl session (40 mins slot) & 1 contributors meetup (friday afternoon) 14:08:41 <acabot> so we will have 1 workroom session per day 14:08:42 <sballe_> +2 14:09:07 <sballe_> it is nice to be a real project now. We get rooms 14:09:09 <acabot> I think its enough compared to Austin summit where it was hard to have everybody on board 14:09:32 <acabot> to focus of 40 min per day on Watcher ;-) 14:09:40 <acabot> s/to/so 14:10:07 <acabot> sballe_ : as I wont be there, I hope you will be able to lead them 14:10:25 <acabot> sballe_ : dtardivel will also help 14:10:45 <sballe_> acabot: I will but will first get there on Tuesday since I have a personal commitemtn in NY the weekend before 14:11:00 <acabot> sballe_ : no pb 14:11:07 <acabot> #info thx to michaelgugino, we now have official OpenStack-Ansible playbooks for Watcher, dtardivel & vincentfrancoise are now core on this repo 14:11:14 <acabot> #link https://github.com/openstack/openstack-ansible-os_watcher 14:11:17 <sballe_> I am staying until Fri late 14:11:47 <gzhai> cool 14:11:58 <acabot> I forgot to mention that the contributors meetup will take place on friday afternoon so please try to stay in Barcelona until Friday EOD ;-) 14:12:10 <sballe_> +2 thx michaelgugino 14:12:31 <acabot> any other announcement ? 14:12:53 <sballe_> We have a small annoucement worung the test system 14:12:58 <sballe_> s/around 14:13:07 <acabot> please do 14:13:18 <sballe_> gzhai: told me that we are getting 15 nodes the first week and 30 the rest of the time 14:13:50 <sballe_> gzhai: has been spending a lot of time preparing but we are confused on the Monasca verusus ceilomter 14:14:02 <sballe_> vincentfrancoise: told me the Monasca BP was on hold 14:14:26 <sballe_> so are we going ot use Ceilomter or Monasca in the tests starting Sept 1st? 14:14:36 <acabot> sballe_ : yes because we were waiting for the Monasca team to discuss Ceilosca 14:14:48 <acabot> we will use Ceilosca 14:15:10 <sballe_> for the Intel scalability tests? 14:15:12 <acabot> the Monasca team did a couple of scalability tests and it should work on our infra 14:15:27 <acabot> yes I mean Intel infra 14:15:45 <sballe_> ok so I am still confused. What do we need inthe Intel Lab? Ceilometer? Monasca? 14:16:05 <sballe_> and is gzhai workingon getting hte right services in our lab 14:16:13 <dtardivel> sballe_: ceilometer + ceilosca + monasca 14:16:34 <dtardivel> monasca will be used as the storage backend 14:16:50 <gzhai> dtardivel: you mean for the test? 14:16:59 <sballe_> dtardivel: so ceilosca is what we are targetting: ceilomter API + Monasca 14:17:03 <vincentfrancoise> sballe_: the way it works is that Ceilometer API ask the backend and the backend is set to be Monasca 14:17:08 <sballe_> dtardivel: yes for the tets 14:17:56 <sballe_> vincentfrancoise: I understand but I just talked to gzhai and he was not installing Ceilosca so I just want to make sure we are all in agreement 14:18:00 <dtardivel> metrics collected by ceilometer agents will be forwarded to Monasca thanks to ceilosca 14:18:32 <sballe_> ok understood. gzhai is that what we are setting up in the Intel Lab? 14:18:38 <vincentfrancoise> sballe_: oh ok, that might an qui proquo then 14:18:40 <gzhai> I assume I use monasca for scalability test, and ceilosca is the future. 14:19:00 <acabot> sballe_ gzhai : as stated https://etherpad.openstack.org/p/watcher-test-environment-specifications we will need Monasca for tests 14:19:15 <jed56> gzhai: no ;-) 14:19:34 <jed56> not exaclty 14:19:46 <jed56> ceilosca is just an abstraction layer 14:19:51 <dtardivel> sballe_: gzhai : Please have a look on https://github.com/b-com/ceilosca. 14:20:17 <gzhai> I know it's same ceilometer API but different backend. 14:20:20 <sballe_> dtardivel: jed56 I totally undertsnad but given that it wasn't spelled out on https://etherpad.openstack.org/p/watcher-test-environment-specifications we now have some confusion 14:20:49 <jed56> sballe_: yes sorry about that 14:20:51 <sballe_> which I just want to get straighten out before we get the system 14:21:07 <dtardivel> gzhai: ceilosca is a ceilometer plugin proposing a new storage backend (monasca) for queries and a new publisher 14:21:10 <sballe_> jed56: no need for an apology ;-) 14:21:43 <sballe_> dtardivel: the scripts you gave gzhai do they deply ceilosca or are they just watcher 14:22:21 <dtardivel> sballe_: do you talk about the mail sent 1 week ago ? 14:22:36 <sballe_> I am sure we have enough time to get this right we just need it done today 14:22:40 <sballe_> dtardivel: yes 14:23:20 <gzhai> It's perfkit in container image. 14:23:21 <acabot> sballe_ : to install ceilosca, please use https://github.com/b-com/ceilosca 14:23:24 <dtardivel> sballe_: no it was about the benchmaker tool we will use in front of OS 14:23:37 <sballe_> dtardivel: ok. 14:24:09 <sballe_> We will update the setup for the test systems and hopefully get the rigth setup by 8/29 14:24:15 <acabot> gzhai sballe_ : is it clear now ? 14:24:40 <sballe_> gzhai: and I will sync-up and yes it is clear now 14:24:42 <dtardivel> gzhai: did you start to deploy Monasca ? 14:25:01 <gzhai> I need test all, ceilometer, ceilosca, and monasca? 14:25:11 <acabot> sorry about that but we where waiting for the Monasca PTL to come back before choosing to use Ceilosca 14:25:21 <acabot> s/where/were 14:25:36 <sballe_> gzhai: no just Ceilosca which is Ceilomter API + Monasca 14:26:05 <acabot> #topic Review Action Items 14:26:11 <acabot> Watcher specs 14:26:14 <tkaczynski> sorry guys, I need to leave earlier today. is there anything related to scoring module which needs my attention? I've published the code for review, applied all the feedback, waiting for merge. in the meantime working on documentation (plugin description, glossary update) 14:26:16 <sballe_> so we need to install Monasca and add the Ceilomter APIs like it is described in https://github.com/b-com/ceilosca 14:26:31 <sballe_> tkaczynski: do you need review of your code? 14:26:47 <sballe_> tkaczynski: I have some time and alsready volunteered to review diga's spec 14:26:54 <acabot> tkaczynski : thx tomasz, I will ask for final reviews of your code 14:27:05 <tkaczynski> sballe_: there were 3 reviews already. but I guess the more the better 14:27:21 <vincentfrancoise> tkaczynski: I think we're ok on your part, I'll come back to you for the doc once you push it for review ;) 14:27:21 <tkaczynski> acabot: thanks! 14:27:26 <sballe_> tkaczynski: I'll give it a look. 14:27:51 <acabot> Update define-the-audit-scope spec has been merged 14:27:58 <tkaczynski> vincentfrancoise: of course, but I will push the doc separately :) 14:27:58 <sballe_> tkaczynski: ditto to vincentfrancoise on the doc 14:27:58 <acabot> #link https://review.openstack.org/#/c/339456/ 14:28:25 <vincentfrancoise> tkaczynski: good 14:28:26 <acabot> alexchadin is working on it right now 14:28:35 <acabot> Graph model describing virtual and physical elements in a data center needs a quick status 14:28:42 <acabot> #link https://review.openstack.org/#/c/298891/ 14:29:01 <acabot> tri2sing : I saw your comments on this spec 14:29:05 <sballe_> acabot: they are working on it. 14:29:38 <acabot> tri2sing : we discussed a lot about it during our last mid-cycle so I would suggest to wait for the code (that should be updated soon) before asking for new PS 14:29:55 <tri2sing> ok 14:29:55 <acabot> s/updated/submitted 14:30:14 <acabot> sballe_ : any timeframe for code delivery N 14:30:15 <acabot> ? 14:30:38 <sballe_> not yet. will ping them 14:30:44 <acabot> ok 14:31:04 <acabot> #action sballe_ ask Intel team about code delivery for https://review.openstack.org/#/c/298891/ 14:31:14 <acabot> Limit concurrent actions invoked by Watcher needs to be retargeted and reviewed 14:31:22 <acabot> #link https://review.openstack.org/#/c/348049/ 14:31:39 <acabot> sballe_ : do you want to review this one as well ? 14:31:50 <sballe_> yes 14:31:57 <acabot> #action acabot sballe_ review https://review.openstack.org/#/c/348049/ 14:32:11 <acabot> Resize instance base on workload needs reviews 14:32:19 <acabot> #link https://review.openstack.org/#/c/359690/ 14:32:45 <acabot> edleafe : about this spec, I saw Nova team discussing "live resize" in the ML, any link with this ? 14:32:49 <sballe_> acabot: I can review this one but I am not really sure I believe Watcher should do that 14:33:20 <acabot> sballe_ : please mention it in your review ;-) 14:33:38 <edleafe> acabot: didn't see that discussion. I have some catching up to do on the ML 14:33:54 <acabot> #action sballe_ acabot jwcroppe hvprash review https://review.openstack.org/#/c/359690/ 14:33:55 <edleafe> acabot: but all resizes are live, aren't they?? 14:34:21 <gzhai> I think resize need reboot. 14:34:28 <gzhai> live resize doesn't. 14:34:36 <acabot> edleafe : it was in the mid-cycle minutes but I didn't understand if they were talking about flavor resinzing... 14:35:10 <edleafe> resize doesn't need reboot 14:35:16 <acabot> #link http://lists.openstack.org/pipermail/openstack-dev/2016-August/100600.html 14:35:21 <edleafe> just a confirmation that it succeeded 14:35:42 <acabot> in "* Next steps for capabilities/policy exposure in the API" 14:35:58 <gzhai> edleafe: how guest can detect the added resources? 14:36:19 <acabot> I will add this ML thread to the review 14:36:43 <acabot> automatic triggering audit needs reviews 14:36:50 <acabot> #link https://review.openstack.org/#/c/342473/ 14:36:51 <edleafe> hmmm... maybe that's resize w/o migration 14:36:57 <edleafe> I'll have to check on that 14:37:05 <acabot> edleafe : thx 14:37:28 <acabot> #action edleafe check is there is a link between nova live resize and https://review.openstack.org/#/c/359690/ 14:37:40 <edleafe> gzhai: using regular sys utilites like free and df? 14:37:58 <diga> acabot: I need your inputs on proposed change & Implementation plan 14:38:18 <acabot> diga : I think we need to add more details in your spec 14:38:22 <diga> jed56: I need your inputs also 14:38:28 <diga> acabot: okay 14:38:36 <acabot> diga : and I dont really see why you rely on Congress... 14:38:56 <jed56> diga: I though that you talked with acabot about that isn't it ? 14:38:57 <diga> ok 14:39:00 <alexchadin> diga: I read through your spec 14:39:06 <alexchadin> hi guys:) 14:39:08 <diga> ok 14:39:13 <vincentfrancoise> alexchadin: hi 14:39:25 <alexchadin> Sorry for being late, work hard on scope 14:39:40 <jed56> hi alexchadin 14:39:57 <diga> what are your thoughts instead of congress 14:40:07 <alexchadin> diga: could you please give us more description about your proposed changes? 14:40:19 <diga> okay 14:40:23 <alexchadin> hi jed56:) 14:40:31 <acabot> diga : you need to justify in your spec why you want to use congress to do that... 14:40:41 <diga> ok 14:40:49 <acabot> and complete your proposed change with more details 14:40:56 <diga> okay 14:41:10 <acabot> #action diga review spec https://review.openstack.org/#/c/342473/ 14:41:13 <alexchadin> diga: IMHO, it should be something like continuous type 14:41:25 <diga> alexchadin: okay 14:41:38 <acabot> diga : yes please look at older specs... 14:41:47 <diga> acabot: sure 14:41:54 <acabot> Watcher 14:42:09 <acabot> as we are facing a code freeze on friday 14:42:18 <acabot> we need to merge 2 things 14:42:24 <acabot> Add Scoring Module implementation needs a final review from core before EOW 14:42:31 <acabot> #link https://review.openstack.org/#/c/351577/ 14:42:44 <acabot> who can do a final review and W+1 ? 14:43:07 <acabot> +2 has already been given by vincentfrancoise 14:43:54 <dtardivel> acabot: I will do it 14:44:04 <sballe_> +2 doe dtardivel 14:44:07 <sballe_> s/for 14:44:16 <acabot> #action dtardivel review https://review.openstack.org/#/c/351577/ 14:44:29 <acabot> Added strategy ID + Action Plan syncing needs final review from core before EOW 14:44:36 <acabot> #link https://review.openstack.org/#/c/325213/ 14:44:53 <acabot> alexchadin : can you do a final review and W+1 this one ? 14:45:11 <acabot> dtardivel already +2 14:45:18 <alexchadin> acabot: I see, will do 14:45:24 <acabot> thx 14:45:40 <acabot> #action alexchadin review https://review.openstack.org/#/c/325213/ before EOW 14:45:52 <acabot> Check unspecified parameters when create audit needs reviews 14:46:00 <acabot> #link https://review.openstack.org/#/c/359818/ 14:46:23 <acabot> actually it needs a new PS 14:46:38 <acabot> #action jinquan1 submit a new PS for https://review.openstack.org/#/c/359818/ 14:46:50 <acabot> #topic Blueprint/Bug Review and Discussion 14:46:59 <edleafe> ok, I did some digging. Nova does shut down the host before resizing; the live resize is meant to avoid having to do that 14:46:59 <acabot> newton-3 is due next week, code freeze on Friday 26th 14:47:19 <acabot> ok 14:47:25 <sballe_> edleafe: thx 14:47:35 <acabot> #link https://launchpad.net/watcher/+milestone/newton-3 14:47:55 <acabot> alexchadin : do you think we will have the audit scope ready ? 14:48:07 <alexchadin> acabot: I work on unit tests 14:48:13 <alexchadin> acabot: I really hope so 14:48:17 <alexchadin> but unit tests... 14:48:42 <acabot> please submit your code on gerrit asap to have 2 full days for review 14:48:43 <alexchadin> I will upload PS today or tomorrow 14:48:48 <vincentfrancoise> alexchadin: can you still push your code so we can start reviewing it in the meatime? 14:48:58 <alexchadin> vincentfrancoise: yeap 14:49:09 <vincentfrancoise> alexchadin: cool :) 14:49:11 <acabot> #action alexchadin submit code for https://blueprints.launchpad.net/watcher/+spec/define-the-audit-scope asap 14:49:15 <dtardivel> alexchadin: with workflow set to -1 14:49:28 <alexchadin> dtardivel: of course:) 14:50:04 <acabot> there are also 2 reviews open for cluster model #link https://review.openstack.org/#/q/status:open+project:openstack/watcher+branch:master+topic:bp/cluster-model-objects-wrapper 14:50:24 <acabot> #info 14:50:24 <acabot> vincentfrancoise added a BP Make Watcher objects versioned as code was already available 14:50:31 <acabot> #link https://blueprints.launchpad.net/watcher/+spec/watcher-versioned-objects 14:50:47 <acabot> Add unit test for watcher/doc.py needs to be discussed 14:50:56 <acabot> #link https://bugs.launchpad.net/watcher/+bug/1527163 14:50:56 <openstack> Launchpad bug 1527163 in watcher "Add unit test for watcher/doc.py" [Low,Triaged] - Assigned to Digambar (digambarpatil15) 14:51:05 <acabot> diga : you want to discuss about this one ? 14:51:14 <diga> yes 14:51:43 <vincentfrancoise> I looked at how they handled doc-related directives in nova and they do not provide any test for them 14:52:15 <diga> Yes, I dont understand how we can write unit test 14:52:19 <vincentfrancoise> but we have to move it out of watcher package and put it in the doc/ folder 14:52:27 <diga> for doc string 14:52:59 <vincentfrancoise> diga: agreed, this would be too much hassle FWIW 14:53:18 <acabot> vincentfrancoise : so the fix would be just moving files ? 14:53:23 <diga> vincentfrancoise: yes, 14:53:49 <vincentfrancoise> acabot: yes, it shouldn't appear in the coverage report for example 14:54:04 <acabot> ok is that clearer diga ? 14:54:08 <diga> acabot: yes, we should follow other project path for doc tests 14:54:10 <diga> yes 14:54:13 <acabot> thx 14:54:17 <acabot> #topic Open discussions 14:54:36 <vincentfrancoise> diga: can you briefly update the bug ticket to reflect what we said ? 14:54:37 <acabot> gzhai : did you try running PKB with docker container ? 14:54:46 <diga> vincentfrancoise: sure 14:54:53 <gzhai> trying. 14:55:02 <gzhai> Some environment issue here. 14:55:09 <acabot> #action diga update big ticket https://bugs.launchpad.net/watcher/+bug/1527163 14:55:09 <openstack> Launchpad bug 1527163 in watcher "Add unit test for watcher/doc.py" [Low,Triaged] - Assigned to Digambar (digambarpatil15) 14:55:21 <sballe_> What's PKB? I have been on vacation for too long and forgot the acronyms 14:55:34 <acabot> PerfKit Benchmarker 14:55:35 <diga> acabot: +1 14:55:45 <acabot> the google tool we want to use to bench watcher 14:55:58 <acabot> #link https://github.com/GoogleCloudPlatform/PerfKitBenchmarker 14:55:59 <sballe_> yes I know what it si ow ;-) 14:56:07 <sballe_> s/is now 14:56:20 <acabot> sballe : ;-) 14:56:31 <acabot> gzhai : any help needed ? 14:56:50 <acabot> gzhai : if you need a VM image instead, please let us knwo 14:56:59 <acabot> s/knwo/know 14:57:06 <gzhai> I think container should work 14:57:14 <acabot> good 14:57:21 <gzhai> will ask david if I can't resolve it. 14:57:29 <acabot> ok 14:57:40 <dtardivel> gzhai: you're welcome :) 14:57:52 <sballe_> gzhai: can we get daily status from you on this? 14:58:02 <gzhai> dtardivel: Thank you! 14:58:12 <gzhai> sballe_: sure 14:58:27 <sballe_> we are getting really close to the 8/29 date and we all need to help to make you successfull 14:58:31 <acabot> we will run internal tests with PKB tomorrow 14:58:32 <gzhai> I have other tasks done now. 14:59:38 <acabot> gzhai : if there is anything we can do to help, just ask on our IRC channel 14:59:56 <acabot> time to close the meeting 15:00:02 <gzhai> acabot: Any scripts for setup, pls. share with me. 15:00:19 <vincentfrancoise> bye everyone 15:00:31 <acabot> #endmeeting