14:01:21 #startmeeting watcher 14:01:22 Meeting started Wed Oct 5 14:01:21 2016 UTC and is due to finish in 60 minutes. The chair is acabot. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:01:23 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:01:25 The meeting name has been set to 'watcher' 14:01:28 morning 14:01:30 o/ 14:01:32 hi 14:01:41 o/ 14:01:42 hi 14:01:43 O/ 14:01:43 hi 14:01:46 agenda for today #link https://wiki.openstack.org/wiki/Watcher_Meeting_Agenda#10.2F04.2F2016 14:02:10 many people there today ! 14:02:22 #topic Announcements 14:02:38 #info please complete topics list in barcelona summit etherpad 14:02:39 Hi 14:02:45 #link https://etherpad.openstack.org/p/watcher-ocata-design-session 14:02:48 this is a good time ;-) next week I won't make it because I will be in california and it will be 2am 14:03:01 ok 14:03:18 we started filling topics with sballe this week 14:03:40 please do it asap as the summit is in 2 weeks now 14:03:56 yes I had Intel people add topics or at least asked them to 14:04:04 hvprash : your new topic is tricky ;-) 14:04:08 I would liek for us to review a draft timeline next week 14:04:19 yeah, have a use case. happy to share :) 14:04:24 pefect 14:05:04 I plan to submit work sessions topics to the foundation by end of next week 14:05:27 so we need to have a final agenda that we will agree for next meeting 14:05:37 acabot: let's sync-up on the work sessions and their agenda on Mon morning? 14:05:55 acabot: +2 14:05:56 sballe_ : you mean monday morning at the summit ? 14:06:10 no next week 14:06:32 ok I plan to submit it next week so no rush 14:06:44 but I want to be sure that all topics will be discussed 14:06:55 agreed 14:07:14 as we have all our BPs for Ocata already defined, we need to work on new strategies and new BPs for the P release 14:07:39 or if we need working sessions for current specs, I also need to know it 14:07:55 we will start the week with the fish bowl session 14:07:57 acabot: I have reviewed Ananth's spec and I would have assumed that he would have submitted it by now 14:08:14 if we want to gather people, we need to find an appealing description 14:08:21 +2 14:08:47 any idea about the topic we should address during this session ? 14:09:29 I dont want to do the Newton retrospective during the fishbowl 14:09:59 what is the goal of a fishbowl session? it is different from normal sessions? 14:09:59 I would prefer an open discussion about possible new strategies enabled by Watcher 14:10:57 "Our traditional largish rooms organized in fishbowl style, with 14:10:57 advertised session content on the summit schedule for increased external 14:10:57 participation. Ideal for when wider feedback is essential." 14:11:05 I think taht is a good idea.. also it would be nice to see if we can consolidate stratgeis so we do not have several that pretty much do the same but a little different 14:11:35 I don;t mena consolidate at this point but more moving forwad decide how we do for example app characterization 14:11:45 ok I will put this in the etherpad but please think about it 14:11:47 +1 14:11:48 and agree on maybe one way 14:12:10 #info Watcher Newton final release has been released 14:12:22 #link https://releases.openstack.org/teams/watcher.html 14:12:32 congrats to all the team 14:12:39 +2 14:12:50 acabot: could we do this during fishbowl: How do we streamline the BPs so we have a more coherent set of strategies moving forward around finerprinting/workload characterization. as well as wokload placement on resources (sballe) 14:12:57 watcher CLI is still waiting for a fix but it will be released soon 14:13:31 Congratulations! 14:13:35 +2 14:13:37 +2 14:13:39 +1 14:13:47 sballe_ : could be interesting but maybe a too much specific if we want to gather people 14:13:51 acabot: We have also a regression about 'parameter' attribute for audit creation, on CLI :/ 14:14:24 I'm investigating if we had again a merge conflict not well resolved .... 14:14:32 dtardivel : ok, should we update the hash as it still possible ? 14:15:01 #link https://review.openstack.org/#/c/378771/ 14:15:08 any other announcement ? 14:15:38 #topic Review Action Items 14:15:46 acabot: not sure if parameter is used a lot for now in audit creation .... It is not critical for me. 14:15:57 dtardivel : ok 14:16:10 a couple of specs needs to be reviewed by cores 14:16:19 #link https://review.openstack.org/#/c/359690/ 14:16:32 #link https://review.openstack.org/#/c/342473/ 14:16:39 #link https://review.openstack.org/#/c/372528/ 14:16:58 I can do that since I am back from cluster hell 14:17:07 #action hvprash sballe_ jwcroppe review https://review.openstack.org/#/c/372528/ https://review.openstack.org/#/c/342473/ https://review.openstack.org/#/c/359690/ 14:17:18 will do 14:17:31 Limit concurrent actions invoked by Watcher needs a new PS 14:17:39 Yeah yeah 14:17:39 #link https://review.openstack.org/#/c/348049/ 14:17:49 Will upload new tomorrow 14:18:15 alexchadin: please add me as a reviewer if I anot already on it 14:18:29 Of course 14:18:31 alexchadin : actually we realized that TaskFlow is already dealing with this in Watcher 14:18:56 jed56 who has implemented TaskFlow can give you more details 14:19:13 hi 14:19:16 Ok 14:19:17 Hi 14:19:43 For more information about taskflow => https://wiki.openstack.org/wiki/TaskFlow 14:20:00 The watcher applier is relying on this library 14:20:14 does TaskFlow handle the limitation of concurent actions ? 14:20:26 How can we limit it then? 14:20:48 IMHO, this is not the responsability of TaskFlow. This is more the scope of the Planner 14:21:03 Task is in charge of executing a flow of actions 14:21:03 I agree 14:21:17 This flow action is build by the watcher planner 14:21:46 ok so in this case we need a spec to define the number of concurrent actions right ? 14:22:08 IMHO, we need a spec to amend the implementation of the default planner 14:22:31 https://github.com/openstack/watcher/blob/master/watcher/decision_engine/planner/default.py 14:22:46 ok so you said that we should look at TaskFlow implementation in more details to write the spec ? 14:23:13 TaskFlow can be just see as a executor 14:23:26 the "brain" is in the planner 14:23:45 So we need to change the implementation of the default planner to add a limit 14:24:10 However, this is not possible for now because we need https://blueprints.launchpad.net/watcher/+spec/planner-storage-action-plan 14:25:14 alexchadin : I would suggest that we put this spec in "stand-by" 14:25:40 as we need a planner improvement before and it will impact your spec 14:25:43 acabot: Agree 14:26:04 ok put it as W-1 in this case 14:26:22 #action alexchadin W-1 https://review.openstack.org/#/c/348049/ 14:26:29 #action alexchadin W-1 https://review.openstack.org/#/c/348049/ 14:26:32 Define grammar for workload characterization needs a new PS 14:26:38 I can try to start work on planner spec 14:26:40 #link https://review.openstack.org/#/c/377100/ 14:26:51 alexchadin : sure you can 14:27:09 I will work with hvprash on Define grammar for workload characterization needs a new PS 14:27:12 acabot, still working on it 14:27:23 alexchadin : https://blueprints.launchpad.net/watcher/+spec/planner-storage-action-plan 14:27:37 sballe_, will do 14:27:39 hvprash : ok do you need feedback ? 14:27:50 Target me then as Assignee 14:28:04 acabot, after the PS :) 14:28:42 #action hvprash sballe_ complete https://review.openstack.org/#/c/377100/ 14:28:51 moving to Watcher 14:29:01 Add Audit Scope Handler needs reviews from core 14:29:08 #link https://review.openstack.org/#/q/status:open++topic:bp/define-the-audit-scope 14:29:59 we need a final review for this 14:29:59 acabot: done 14:30:03 ok 14:30:11 #action dtardivel review https://review.openstack.org/#/q/status:open++topic:bp/define-the-audit-scope 14:30:21 Graph cluster model implementation needs a new PS 14:30:28 #link https://review.openstack.org/#/c/362730/ 14:30:43 sballe_ : can you ask Kevin to update his code ? 14:31:20 any other reviewer is welcome on this 14:31:53 Add service supervisor needs reviews 14:31:59 #link https://review.openstack.org/#/q/status:open+project:openstack/watcher+branch:master+topic:bp/watcher-service-list 14:32:42 licanwei : as you have reviewed doc, could you review the code ? 14:33:07 acabot: ok 14:33:18 #action licanwei review https://review.openstack.org/#/c/371632/ 14:33:19 thx 14:33:34 #action sballe_ add a new PS for https://review.openstack.org/#/c/362730/ 14:33:47 Added Tempest API tests for /scoring_engines needs reviews 14:33:54 #link https://review.openstack.org/#/c/371558/ 14:34:12 sorry it needs a new PS 14:34:25 #action vincentfrancoise add a new PS for https://review.openstack.org/#/c/371558/ 14:34:38 Watcher versioned objects needs reviews 14:34:42 acabot: ok will do by EOD 14:34:46 #link https://review.openstack.org/#/q/status:open+project:openstack/watcher+branch:master+topic:bp/watcher-versioned-objects 14:34:52 acabot: +1 14:34:53 acabot: yes 14:35:11 #action dtardivel review https://review.openstack.org/#/q/status:open+project:openstack/watcher+branch:master+topic:bp/watcher-versioned-objects 14:35:27 I'd like to have more reviewers on this as its a large set of code 14:35:48 +1 14:35:55 reviewers and testers neeeded ;-) 14:36:03 #action licanwei review https://review.openstack.org/#/q/status:open+project:openstack/watcher+branch:master+topic:bp/watcher-versioned-objects 14:36:23 Will review some of them 14:36:37 #action alexchadin review https://review.openstack.org/#/q/status:open+project:openstack/watcher+branch:master+topic:bp/watcher-versioned-objects 14:36:43 alexchadin licanwei: thx 14:37:16 Watcher CLI 14:37:22 Add service support needs reviews 14:37:28 #link https://review.openstack.org/#/c/373533/ 14:38:19 +1, but waiting for new PS on the watcher side 14:38:30 ok 14:38:42 #action dtardivel review https://review.openstack.org/#/c/373533/ 14:38:57 #topic Blueprint/Bug Review and Discussion 14:39:07 Improve the storage of the ActionPlan has many dependencies and should be started asap 14:39:16 #link https://blueprints.launchpad.net/watcher/+spec/planner-storage-action-plan 14:39:37 alexchadin : I have assigned it to you, could you start the spec ? 14:39:51 Yes 14:39:56 ok thx 14:40:10 #info we still have many unassigned blueprints for Ocata 14:40:18 #link https://blueprints.launchpad.net/watcher/ocata 14:40:43 licanwei : did you take one or not ? 14:41:01 acabot: yes 14:41:33 licanwei : which one ? 14:41:56 any one :) 14:42:03 sballe_ : could you ask Ananth to take https://blueprints.launchpad.net/watcher/+spec/context-aware-workload-mgmt-using-fingerprints ? 14:42:24 licanwei : it depends your interests :-) 14:43:01 sorry, i would have taken but already 2 in my plate 14:43:12 licanwei : https://blueprints.launchpad.net/watcher/+spec/stale-action-plan ? 14:43:17 take it where? 14:43:48 sballe_ : in launchpad so that we have an assignee for it 14:43:52 you mean to assign it to himself 14:43:58 yes 14:44:03 yes I will do that 14:44:08 thx 14:44:39 acabot: ok 14:45:08 licanwei : ok could you start with a spec ? 14:46:15 we also have open bugs #link https://bugs.launchpad.net/watcher 14:46:33 acabot: yes ,i can 14:46:44 licanwei : thx 14:46:49 #topic Open discussions 14:47:19 as already said last week, we need more reviewers 14:47:30 droveI plan to review again 14:47:41 droveI ? 14:48:07 gzhai doesn't do any core review 14:48:11 s/I 14:48:24 irc did some sort of text completion. 14:48:29 :-) 14:48:40 I think we need new cores from IBM & Intel 14:48:49 if we want to keep the project on track 14:49:09 hvprash : do you plan to have a core reviewer on Watcher code ? 14:49:19 hvprash : I mean on Walmart side 14:49:35 acabot, i will have an answer next week and some updates 14:49:47 hvprash : ok 14:50:20 sballe_ : on your side, any option to have a new core ? 14:51:20 I need to look at it. let me see what i can figure out 14:52:12 ok 14:52:20 any other topic to discuss ? 14:52:47 did you plan your parties in Barca ? :-) 14:52:58 acabot: fyi, I will push a new PS about strategy template doc, in 10mn :) 14:53:09 dtardivel : ok cool 14:53:10 acabot, is it the same etherpad https://etherpad.openstack.org/p/watcher-ocata-design-session to discuss general watcher topics in barcelona ? 14:53:24 so all, please review it ;) 14:53:49 hvprash : yes I try to avoid having multiple etherpads 14:53:59 hvprash : do you need a specific one ? 14:54:04 ok thx, will keep updating then :) 14:54:36 hvprash : thx 14:54:47 have some topics. created one under general topics in the etherpad 14:55:05 hvprash : I can see that, thx 14:55:19 I can close the meeting 14:55:23 thank you guys 14:55:29 thanks everyone ! 14:55:31 bye 14:55:35 bye 14:55:37 bye 14:55:40 bye 14:55:41 #endmeeting