14:00:15 <jed56> #startmeeting watcher 14:00:16 <openstack> Meeting started Wed Apr 6 14:00:15 2016 UTC and is due to finish in 60 minutes. The chair is jed56. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:17 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:19 <openstack> The meeting name has been set to 'watcher' 14:00:28 <jed56> hi everyone 14:00:29 <sballe_> o/ 14:00:33 <AndChat|176409> Hi 14:00:40 <tkaczynski> hi 14:00:47 <jed56> agenda for today #link https://wiki.openstack.org/wiki/Watcher_Meeting_Agenda#04.2F06.2F2016 14:01:04 <vincentfrancoise> o/ 14:01:17 <vtech> o/ 14:01:33 <jed56> #topic Announcements 14:01:36 <tkaczynski> is jed56 the new acabot? :) 14:01:51 <jed56> tkaczynski: acabot is in holidays :) 14:01:56 <alexchadin> Hi 14:02:09 <jed56> #info we released Watcher mitika-rc-final yesterday with some delay with official schedule. 14:02:21 <jed56> The April 1st was finally a joke ;) 14:02:21 <sballe_> great job everybody 14:02:23 <alexchadin> Great! 14:02:31 <jed56> The RC-final details are available at #link https://launchpad.net/watcher/+milestone/mitaka-rc-final 14:02:40 <jed56> There is three new watcher packages 14:02:54 <jed56> watcher v0.26.0 python-watcherclient v0.22.0 and finally watcher-dashboard v0.2.0 14:03:08 <jed56> The documentation of watcher has been updated (0.26.0) #link https://factory.b-com.com/www/watcher/doc/watcher/index.html 14:03:21 <jed56> thanks everyboy! 14:03:26 <jed56> everybody 14:03:40 <jed56> any other announcement ? 14:04:51 <sballe_> the slides for the summit are close to being done. I have an Intel person review the slides next week 14:05:14 <jed56> sballe_: great! you share again the link please 14:05:18 <jed56> can 14:05:25 <sballe_> let me find it 14:05:30 <jed56> thanks 14:05:46 <sballe_> Here it is: https://docs.google.com/presentation/d/1pGaimFOkpZ66bWH8wWkc5eZLvxFovUj2V3YfZsQMKHE/edit#slide=id.p4 14:06:11 <tkaczynski> it would be good to share some agenda for the summit, especially in context of watcher 14:06:19 <tkaczynski> where/when do we meet etc. 14:07:11 <jed56> tkaczynski: yes it is a good idea, can we organize when acabot will return ? 14:07:37 <sballe_> so my understanding is that we were going ot meet on Tuesday afternoon. But I agree let's have acabot organize that agenda 14:07:49 <tkaczynski> of course, as long as this happens before the summit 14:07:58 <jed56> tkaczynski: yes of course :) 14:08:01 <jed56> #topic Review Action Items 14:08:21 <jed56> There are many specifications which have been reviewed several times. IMHO, we can merge some of them to start the implementation. 14:08:29 <jed56> I would like that these specifications are merged as soon as possible: 14:08:36 <jed56> Watcher Scoring Module #link https://review.openstack.org/#/c/289880/ 14:08:39 <jed56> Provide efficacy indicators #link https://review.openstack.org/#/c/283449/ 14:08:44 <jed56> Add Overload standard deviation #link https://review.openstack.org/#/c/286153/ 14:08:50 <tkaczynski> jed56 +2 ! 14:08:53 <jed56> Who could take some time to look at these reviews in priority, please ? 14:09:05 <sballe_> I will review the scoring engine today 14:09:40 <jed56> sballe_: awesome! 14:09:58 <tkaczynski> I can have a look at the other two if my limited knowledge helps 14:10:02 <sballe_> it takes a long to to review because it is big so I wont; get to the next on euntil tomorrow 14:10:32 <jed56> #action sballe review #link https://review.openstack.org/#/c/289880/ 14:10:34 <vincentfrancoise> I'll try and review some of them as well 14:10:52 <gzhai2> I can review Provide efficacy indicators 14:11:04 <jed56> #action tkaczynski reviews https://review.openstack.org/#/c/283449/ and https://review.openstack.org/#/c/286153/ 14:11:17 <jed56> #action ghzai2 review https://review.openstack.org/#/c/283449/ 14:11:29 <jed56> We also need to have more more reviews for on Airflow event based migration strategy #link https://review.openstack.org/#/c/275474/ 14:12:08 <jed56> tpeoplesvacation: When you plan to send a new patchset for WIP: Cluster model objects wrapper https://review.openstack.org/#/c/287019/ 14:12:16 <sballe_> I will look at at: https://review.openstack.org/#/c/275474/ 14:12:28 <jed56> #action tpeoplesvacation send a new patchset WIP: Cluster model objects wrapper https://review.openstack.org/#/c/287019/ 14:12:49 <jed56> #action sballe review https://review.openstack.org/#/c/275474/ 14:13:01 <jed56> We have two new specs from Grigorios Katsaros that need first reviews 14:13:09 <jed56> Graph model describing virtual and physical elements in a data center 14:13:09 <jed56> #link https://review.openstack.org/#/c/298891/ 14:13:09 <jed56> Consolidation of multiple Action-plans into a single one 14:13:09 <jed56> #link https://review.openstack.org/#/c/298871/ 14:13:20 <sballe_> jed56: are these for now or later release? 14:13:44 <jed56> later release 14:13:55 <jed56> i think 14:14:00 <jed56> but this is my opinion 14:14:18 <sballe_> I'll give link https://review.openstack.org/#/c/298891/ a quick loook but want to focus on things we want to get merged ASAP 14:14:35 <sballe_> jed56: I agree with you on the later 14:15:07 <jed56> sballe_: yes we should focus our effort 14:15:21 <jed56> Since we have released the watcher mitika final the specs are open for Newton 14:15:26 <jed56> #action acabot sballe jwcroppe review #link https://review.openstack.org/#/c/301753/ 14:15:40 <jed56> #topic Blueprint/Bug 14:16:25 <alexchad_> i've uploaded bug fix 14:16:26 <alexchad_> https://review.openstack.org/#/c/301190/ 14:16:26 <jed56> Many bugs have been merged in mitika-rc-final thanks to everybody! 14:17:05 <jed56> alexchad_: thanks i will take a look 14:17:17 <alexchadin> :) 14:17:24 <jed56> We realized that the blueprint get-goal-from-strategy have some lacks in the specification. 14:17:47 <vincentfrancoise> alexchadin: I gave it a +2 14:17:50 <jed56> We will push a new patchset in the watcher-specs asap to tackle the issue of goal syncing and watcher DB schema for goal (the uuid and id problems). 14:17:54 * edleafe is fighting network connectivity today... 14:18:20 <jed56> We noticed that the attribute id is used by oslo.db to handle soft delete feature. 14:18:45 <alexchadin> and this one fix is coming very soon: https://bugs.launchpad.net/watcher/+bug/1532843 14:18:47 <openstack> Launchpad bug 1532843 in watcher "Can create Audit with invalid parameters" [Low,Triaged] - Assigned to Alexander Chadin (joker946) 14:19:55 <jed56> #link https://review.openstack.org/#/c/301774/ 14:20:22 <jed56> this is a WiP 14:21:01 <jed56> gzhai2 tpeoplesvacation we will add uuid and id https://review.openstack.org/#/c/301774/8/doc/source/images/get_goal_from_strategy_class_diagram.png,unified 14:21:14 <vincentfrancoise> gzhai2: In this, we add back the ID and UUID just like tpeoplesvacation was requesting + clarified how we sync goals and strategies with the Watcher DB 14:21:44 <gzhai2> vincentfrancoise: Great! 14:21:46 <jed56> gzhai2: thanks for your reviews 14:22:28 <vincentfrancoise> I actually started to work on making these changes 14:23:04 <vincentfrancoise> So you can work on your BP asap 14:23:40 <gzhai2> vincentfrancoise: My BP? 14:25:18 <vincentfrancoise> gzhai2: yes https://blueprints.launchpad.net/watcher/+spec/optimization-threshold 14:26:00 <gzhai2> vincentfrancoise: you mean I rebased my patch against your patch set？ 14:26:01 <vincentfrancoise> gzhai2: because your BP has a dependency on mine 14:26:10 <vincentfrancoise> on yours* 14:26:36 <gzhai2> ok. you mean change the spec accroding to your spec, right? 14:26:49 <vincentfrancoise> gzhai2: when I will be done, yes 14:27:07 <vincentfrancoise> gzhai2: no just your code I mean 14:27:26 <gzhai2> got it. 14:27:33 <jed56> any other bugs or blueprints ? 14:29:34 <jed56> okay let's continue 14:29:38 <jed56> #topic open discussion 14:30:08 <jed56> We have to postpone some blueprints and bugs from mitaka to newton. 14:30:13 <jed56> Do we need to retarget blueprints from Mitaka to Newton in watcher-specs ? 14:30:47 <jed56> IMHO, every can move his own blueprint to newton 14:30:55 <jed56> *everybody 14:31:02 <sballe_> +1 14:31:30 <gzhai2> ok 14:31:30 <jed56> you can rebase your patchset to https://review.openstack.org/#/c/301753/ 14:31:53 <jed56> or you can wait that this patchset is merge 14:32:35 <jed56> #action acabot sballe jwcroppe review #link https://review.openstack.org/#/c/301753/ 14:32:53 <jed56> Before the Openstack Summit in AUSTIN, we should run a battery of tests on watcher to be sure that everything is working (Watcher and the available strategies). 14:32:57 <jed56> Who could take some time to deploy on test watcher on their own infrastructure ? 14:34:01 <vincentfrancoise> I actually test every single changeset against a multinode devstack 14:34:16 <jed56> vincentfrancoise: did you try every strategies ? 14:34:42 <vincentfrancoise> Yes apart from the outlet temp one since I don't have the IPMI metrics 14:34:44 <gzhai2> I can cover outlet temperature strategy 14:35:06 <sballe_> gzhai2: +1 14:35:06 <jed56> #action gzhai2 test watcher with outlet temperature strategy 14:35:12 <alexchadin> I try strategies sometime 14:35:17 <gzhai2> vincentfrancoise: You need one Node Manager machine:) 14:35:19 <vincentfrancoise> for the vm_workload_conslidation, I just ran it once 14:35:42 <jed56> What is the strategy we wants to show at Openstack Summit ? 14:35:45 <alexchadin> I have test cluster with 4 nodes 14:36:04 <vincentfrancoise> only 2 nodes on my side 14:36:08 <jed56> alexchadin: great! how many VMs ? 14:36:32 <alexchadin> now about 10, can raise up to 100 14:36:49 <vincentfrancoise> I am currently waiting for https://review.openstack.org/#/c/299809/ to be merged so we can have mutinode tempest tests on the gate 14:36:51 <alexchadin> it's just work env 14:37:53 <jed56> #action acabot organize the agenda for the Openstack Summit (where/when/) 14:38:51 <jed56> someone wants to add something ? 14:39:59 <gzhai2> who can review https://review.openstack.org/#/c/292188/ 14:40:46 <jed56> #action jed56 vincentfrancoise https://review.openstack.org/#/c/292188/ 14:41:01 <gzhai2> thanks! 14:42:04 <tkaczynski> #endmeeting :) 14:42:16 <jed56> yes i think so 14:42:16 <jed56> :) 14:42:19 <jed56> bye 14:42:20 <jed56> #endmeeting